Datasets:

ianiket23

/

podcast_llama_chat_format-5k

like 0

<s>[INST] What advice would you give to those young people today? Maybe in high school and college at Stanford, maybe to your grandchildren? About how to, uh, have a career they can be proud of, Have a life they can [/INST] be proud of, pursue careers that are in the public interest, you know, in one fashion or another. And not just in their interest. Um, and and that would be I mean, it's not bad to pursue a career in your own interest. I mean, as long as you're it's something that's useful and positive for the, you know, for for their families or whatever, but yeah, so I mean, I I try to try to advise kids to find themselves somehow, you know, find out who they want to be and what they want to be and try to pursue it. And the NGO world is growing, as you know, and, uh, and a lot of young people are kind of throwing themselves into it and and, yeah, Human Rights Watch and that kind of stuff, and you know, they want to do that kind of work, and it's it's very admirable. </s><s>[INST] I tend to think that even if you're not working in human rights, there's a certain way in which, if you live with integrity, the I believe that all of us are. Many of us have a bunch of moments in our lives when we're posed with the decision. It's a quiet one. Maybe you'll never be written about or talked about when you get to choose whether you, uh, there's a choice, that it's difficult to make. It might require sacrifice, but it's the choice that the best version of that person would make. Um, that's the best way I can Sort of say how to act with integrity is the very thing that would resist the early days in Nazi Germany. It sounds dramatic to say, but those little actions, um, and I feel like the best you can do to avoid genocide on scale. It's for all of us to live in that way to within those moments, unrelated potentially to human rights to anything else is to take those actions like, I believe, that all of us know the right thing to do. [/INST] You know? That's right. I think that's right. You put it very well. I couldn't have done it better myself. No, no, I I agree. I agree completely that there are, you know, you know, to live, to live with truth, which is what possible hovel used to say, This famous Czech dissident, you know, talked about living in truth, but also to live with integrity. Um, and that's really super important. </s><s>[INST] Well, let me ask you about love. What role does love playing this whole thing in the in the human condition and all of the study of genocide, it does seem that hardship in moments brings out the best in human nature, and the best in human nature is expressed the love. [/INST] Well, as I already mentioned to you, I think hardship it can is not a good thing for, you know, it's not the best thing for love. I mean, it's better. It's better to not have to suffer and not have to. Yes, I think it is. I think it's, you know, as I mentioned to you, you know, studying concentration camps. You know, this is not a place for love. It happens. It happens. But it's not really a place for love. It's a place for rape. It's a place for torture. It's a place for killing in this place for in human action, one to another, you know, and also, as I said, among those who are suffering not just between those who are and and then their whole gradations. You know the same thing in the gulag. You know, there are gradations all the way from the criminal prisoners who beat the hell out of the political prisoners. You know who then have others below them? Who? They beat them, you know, So everybody is being the hell out of everybody else. So I would I would not idealize in any way suffering as a you know, uh, a source of source of beauty and love. I wouldn't do that. I think it's a whole lot better for people to be relatively prosperous. I'm not saying super prosperous, but to be able to feed themselves and to be able to feed their families and how's their families and take care of themselves, Um, you know, to foster loving relations between people. And, you know, I think it's no accident that, um, you know, poor families have much, you know, worse records when it comes to crime and things like that, you know, and and also to wife beating and to child abuse and stuff like that. I mean, you just you don't you don't want to be um, poor and indigent and not have a roof over your head be homeless. I mean, it doesn't mean again, you know, homeless people are mean people. That's not what I'm trying to say. What I'm trying to say is that you know what we want to try to foster in this country and around the world and one of the reasons, you know? I mean, I'm very critical of the Chinese in a lot of ways, but, I mean, we have to remember they pull the country out of horrible poverty, right? And I mean, there's still poor people in the countryside. There's still problems, you know, with with want and need among the Chinese people. But, you know, there were millions and millions of Chinese who were living at the bare minimum of life, which is no way to live, you know, and no way again to foster love and compassion and getting along. So I want to be clear. I don't speak for history, right? I'm giving you. I mean, it used to be historians, you know, in the 19th century, who really thought they were speaking for history. You know, I don't think that way at all. I mean, I understand I'm a subjective human being with my own my own points of view in my own opinions. </s>

<s>[INST] And it changes the way you interact in future disagreements. She has those moments. [/INST] Yeah, you can't unlearn. Now you know this about your partner, </s><s>[INST] you know what they're sensitive to. Again, you kind of see the beauty and the flaws then. Right? It all makes sense, right? Yeah, it all kinda makes sense. [/INST] Yeah, right. It all makes sense. Yeah, it all kinda makes sense. So you maybe were in this dumpster dive in your head of how your partner sucks and all the things that are wrong with them and it's so hopeless. And then you get this light shining through and you realize, oh my God, of course they would be sensitive to that. And suddenly it's not about all the ways your partner is wrong and proving that they're wrong. It's just, how can I in the future make sure they do not feel this again? I would never want this person I love </s><s>[INST] to misunderstand me and feel so unloved. What are you, the early days of that? What do you think about the whole dating, modern dating process? How do you find a partner that you can stay with [/INST] for the rest of your life? So we are absolutely doing it wrong. And, but there is a way you can do it. And I am such a fan of the psychologist, Ty Tashiro. I adore him. He's brilliant. He's lovely. He's also very humble. Just a wonderful salt of the earth guy. I'm going to tell you a very true story here. Okay. I was in a bad relationship and I was at a psychology conference with my partner at the time. We were both at this conference and we were sitting in a lecture hall there for Ty Tashiro to do his talk that day on his phenomenal research on relationship satisfaction and dating. And I'm sitting next to him and we'd been, you know it was just always unpleasant on trips. There were always fights. We're sitting there and Ty Tashiro starts talking about his research and how he found that most people are, you know signing this agreement, getting married and doing it based on the love endorphins. And really only about 35% of anybody who's married is actually happy. And he said, so then, you know, and It's a pretty low number. Exactly. But here's what I love about Ty Tashiro is he didn't stop there. He wanted to know what those people who were happy had in common. And then same thing with the people who were unhappy. He found a couple fascinating patterns. So the couples who were happy tended to rate their partners higher in three different traits. And I love talking about this because if you are somebody who can follow instructions you can find this, I mean, very easily. Those three traits tend to be conscientiousness. Okay. And I love the word conscientiousness because it's not just kindness. Kindness is a good way to think of it but you can be kind and kind of be a pushover and that's not attractive. Conscientiousness is smart, attentive. It's somebody who reads into a text message and thinks, wow, she was making herself very vulnerable there. </s><s>[INST] That's conscientiousness. [/INST] I like how you're just doing a compliment. It's true. It's a certain intelligence, awareness and attunement. And then on top of that, conscientiousness is motivated. So you can't be on your ass all day and be conscientious because then you can't meet the needs that you anticipate about the person. So conscientious is that guy who drives the car around in the rainstorm so his wife's hair doesn't get met. It's my husband who checks my alarm for me every morning because he knows I'm terrible at time management and he makes sure that I set it a reasonable amount of time before my first meeting and not the 20 minutes I think I need. And then he'll come wake me up with a cup of coffee. That is ultimate conscientiousness. And it is true. I mean, I will tell you as somebody who's with a conscientious partner, your love increases over time as you continue to feel grateful and admiring of that person. The second one, you want somebody who is low in a big five personality trait called neuroticism. You want somebody emotionally stable in a way. Now this doesn't mean you can't have somebody who doesn't get the blues or struggle with mental health issues. Trust me, Ty is with somebody who I'm all over the place. But you want somebody who kind of owns their shit and isn't going to just be emotionally unstable all over. You want somebody who is generally happy and has some life satisfaction. Having a partner who has serious, not mental health issues, but unmitigated emotional distress and instability is really hard on the partner. And it's really hard on other family members, including children if you have children. So it's just a predictor of happiness. </s><s>[INST] So there's a certain threshold of chaos that if you exceed it, it's going to be destructive [/INST] to a long-term relationship. Exceed it is going to a long-term relationship. A perfect description about chaos. Not the mystery chaos you love with your little poet brain. I'm talking more like just somebody who, there's just no peace. There's no peace. There's a problem with everything. Everything becomes more difficult. Going to a party is a chore. You don't know if they're gonna have a meltdown at the party or how many complaints about your friends or everything is a problem. So you want somebody who has just some resiliency, I think is a good term for it. </s><s>[INST] Some flexibility. But then again, yeah, spice is okay, but not too much. [/INST] Right, right. Flexibility, resiliency, easy going. Yeah. Okay. The third is really interesting, I think. So he found that having a partner with sort of moderate adventurousness, not high adventurousness actually leads to greater satisfaction. And the reason for that is high adventurousness equals novelty seeking, shiny new things. And so if you're in a monogamous relationship, if that is what's important to you, it's going to be very hard for a partner who is novelty seeking to be faithful. So that will cause a lot of pain. But also novelty seeking people tend to always have new projects, new interesting things. And so their attention is drawn away from the relationship. </s>

<s>[INST] the same as our TikTok, right? That's what I mean. Which aspect of it is not the same? [/INST] Well, I just think, so what is TikTok? TikTok is social media where people upload videos, right, </s><s>[INST] of silly videos. So I guess there might be- Well, there's humor, there's attention, there's ability to process, there's ability for intelligent organisms to collaborate on ideas and find humor in ideas and play with those ideas, make them viral memes. Humor seems to be kind of fundamental to the human experience. [/INST] And I think that that's a really interesting question we can ask. Is humor a fundamental thing in the universe? I think maybe it will be, right? In terms of you think about in a game theoretic sense, the emergence of humor serves a role in our game engine. And so if selection </s><s>[INST] is fundamental in the universe, and so is humor. Well, I actually don't know exactly what role humor serves. Maybe it's from a chemical perspective, it's like a catalyst for, I guess it serves several purposes. One is the catalyst for spreading ideas on the internet, that's modern humor. But humor is also a good way to deal with the difficulty of life. It's a kind of valve, release valve for suffering. Throughout human history, life has been really hard. And for the people that I've known in my life who've looked through some really difficult things, humor is part of how they deal with that. It's usually dark humor. But yeah, it's interesting. I don't know exactly what's the more mathematically general way to formulate what the hell is humor? What humor does it serve? But I still, we're kind of joking here, but it's a counterintuitive idea to me to think that life elsewhere in the universe is very different than life on Earth. And also, each instantiation of life is likely very different from each other. Maybe there's a few clusters of similar life, but it's much more likely is what you're saying. To me, it's a kind of novel thought, I'm not sure what to do with it. But you're saying that it's more common to be a weird odd cast in the full spectrum of life than it is to be in some usual cluster. So every instantiation of a kind of chemistry that results in complexity that's autonomous and self-replicating, however the hell you define life, that is going to be very different every time. I don't know. It feels like a selection is a fundamental kind of directed force in the universe. Won't selection result in a few pockets of interesting complexities? If we ran Earth over again, over and over and over, you're saying it's going to come up with, [/INST] there's not going to be elephants every time? Yeah, I don't think so. I think that there will be similarities. And I think we don't know enough about how selection globally works. But it might be that the emergence of elephants was wired into the history of Earth in some way, like the gravitational force, how evolution was going, the Cambrian explosions, blah, blah, blah, the emergence of mammals. But I just don't know enough about the contingency, the variability. All I do know is you count the number of bits of information required to make an elephant and think about the causal chain that provide the lineage of elephants going all the way back to </s><s>[INST] Luca. There's a huge scope for divergence. Yeah, but just like you said, with chemistry and selection, the things that result in self-replicating chemistry and self-replicating organisms, those are extremely unlikely, as you're saying. But once they're successful, they multiply. So it might be a tiny subset of all things that are possible in the universe, chemically speaking, it might be a very tiny subset is actually successful at creating elephants, [/INST] or elephant-like, or slash human-like creatures. Well, there's two different questions. The first </s><s>[INST] one, if we were to reset Earth and to start again- At the different phases, sorry to keep interrupting. [/INST] Yeah, no. If we restart Earth and start again, say we could go back to the beginning and do the experiment or have a number of Earths, how similar would biology be? I would say that there would be broad similarities, but the emergence of mammals is not a given unless we're going to throw an asteroid at each planet each time and try and faithfully reproduce what happened. Then there's the other thing about when you go to another Earth-like planet elsewhere, maybe there's a different ratio of particular elements, maybe the bombardment at the beginning of the planet was quicker or longer than Earth. I just don't have enough information there. What I do know is that the complexity of the story of life on Earth gives us lots of scope for variation, and I just don't think it's a reasonable mathematical assumption to think that life on Earth that happened again would be the same as what we have now. </s><s>[INST] Okay, but you've also extended that to say that we might, as an explanation for the Fermi paradox, that that means we're not able to interact with them. That's an explanation for why we haven't at scale heard from aliens is they're different than us. [/INST] Well, right now, we've only been looking for, say, 70, 80 years, so I think that the reason we have not found aliens yet is that we haven't worked out what life is. </s>

<s>[INST] what do you think is the connection john between this ego pride thing martial arts and actual violence in our, with our ancestors, do you think? Do you ever plug into that? You think there's echoes of something going on there? Or like you mentioned, you have flaws and demons. Is it deep in there somewhere? Do you think we're struggling with those demons? Yeah. You'll need to patch up your question a little bit though, it's going in several different directions, wow, that was not only my being dominant, dominant, interviewing questions, it's [/INST] like, okay, now we interview you, but you went down the evolution. </s>

<s>[INST] Well, basically, can we step back? So what was the light in the image taken in 2019? [/INST] Well, not taken in 2019, presented in 2019. So here's the puzzle. What they really saw, all right. So the black holes tend to gather stuff that swirls around it and they don't know what that stuff is made of, they don't know what its temperature is. They don't know what kind of magnetic fields there are around there. So the form of the image has a lot of unknowns in it that it's dependent on many other things other than the geometry of the black hole. So most of what you're learning is about the stuff. Now the stuff, the swirling stuff, the hot swirling stuff is interesting as hell, but it's not as interesting as the black hole which are the most, in my view, the most interesting things in the universe. So you don't want to just learn about the stuff, you want to learn about the black hole that it's swirling around. </s><s>[INST] So at the very first step, at the very primitive level, this is just a big lead for human civilization to be able to see a black hole and the way you can see it is because there's stuff around it. But you don't get to learn much about the black hole, but you get to learn more about [/INST] the stuff just from the image. Yeah. But you're not gonna learn about the details before you've </s><s>[INST] even seen it. Cuz there's too many parameters, it's too many variables that govern the stuff. [/INST] Yeah. So then we found a very wonderful way to learn about the black hole. And here's how it works. A black hole is a mirror. And the way it's a mirror is if light, a photon, bounces off your face towards the black hole and it goes straight to the black hole, just falls in, you never see it again. But if it just misses the black hole, it'll swing around the back and come back to you and you see yourself from the photon that went around the back of the black hole. But not only can that happen, the black hole, the photon can swing around twice and come back. So you actually </s><s>[INST] see an infinite number of copies of yourself. Like with a little bit of a delay. [/INST] With a little bit of a delay, right. This is awesome. Yeah. And in fact, </s><s>[INST] I mean, we're not used to an object that bends light like that, right? Yeah, yeah. [/INST] So you're going to get some trippy effects. Yeah. And in fact, one of my students has made a really awesome computer animation of this, which I'm going to show at a public lecture in a couple of weeks where the audience will see infinitely many copies of themselves swirling around the black hole. So a black hole is like a hall of mirrors, like in an apartment store where you go and there's the three mirrors and you see infinitely many copies of yourself. Think of the black hole as the mirror, and you go in there with your clothes. If you want to know about your clothes, you just look at the direct image. You're not learning anything about the configuration of mirrors. But the relation of the image you see in front of you to the one you see at the side and the next one and so on depends only on the mirrors. It doesn't matter what clothes you're wearing. So you can go there a thousand times wearing different clothes, but each time there will be the same relation between the subsequent images. And that is how we're going to learn about the black holes. We're going to take the stuff that is swirling around and we're going to tease out the subsequent images and look at their relation. And there's some very beautiful, really beautiful mathematics, which we were surprised to realize with the volumes and volumes of papers on black holes and their properties this particular... Because it was a physical </s><s>[INST] question that had never been asked in exactly this way. So basically, you're looking at the... The relation between the subsequent images. But those are ultimately formed by photons that are swirling around the... Photons that are orbiting. So the photon ring are the photons that orbit around. And beyond. So like orbit and lose orbit. Like are they... Yeah. Wow. And that starts to give you, what can you possibly figure out mathematically about the [/INST] black hole? The geometry of it, the spin of it. The geometry, the spin, and you can verify things behaving. We have never seen a region of space time with such high curvature. I mean, the region around a black hole is crazy. It's not like in this room. The curvature is everything. </s><s>[INST] You spent probably enough time with the math and the photons. Can you put yourself in that space? So we're having a conversation in pretty peaceful, comfortable, flat space. Are you able to put yourself in a place around a black hole? [/INST] Yeah, I'm able to imagine that kind of thing. Yeah. So for example, and actually, there's a wonderful movie, Interstellar. And in that movie, Kip Thorne, of course, is a great theoretical physicist, experimental, who later won the Nobel Prize for LIGO. And that movie is very accurate, scientifically. And there's some funny statements in there that of the hundred million people who saw that movie, there can't be more than 10 or 20 understood about why Matthew McConaughey is ejecting the trash in a certain direction in order to... But for example, if I were a spinning black hole right here, if I were spinning fast enough, you wouldn't be able to stay still there. You'd have to be orbiting around like that. You'd have </s><s>[INST] to have your microphone on a... But I wonder what the actual experience is, [/INST] because I mean, space itself is curved. Well, if space gets very curved, you get crushed. Body gets ripped apart because the forces are different on different parts. Sure. Okay, so that would be... But if it can be less curved so that the curvature is very noticeable, </s>

<s>[INST] but that's with of course I would not, I made that up completely. It's it's all just like my girlfriend in Canada, she's completely a figment of my imagination. Nevertheless, Winston Churchill um is somebody, I mean, just looking at reading the rise and fall of the Third Reich is an incredible figure that um that to me, so much of World War Two is marked leading up to the wars marked by stunning amounts of cowardice by political leaders. And it's fascinating to watch here. This person, clearly with the drinking and a smoking problem. [/INST] I don't understand why that's a negative. </s><s>[INST] No, I didn't say [/INST] you throw it in as if it is. </s><s>[INST] No, well, it's called humor. I'll explain it to you one day what that means, [/INST] but he still explained dry him. He </s><s>[INST] stood up. He he stood up to what we now see as evil when at the time it was not so obvious to see. Um, you know, so that that's that's just a fascinating figure of Western civilization. I'd love to get your comments. [/INST] The real criticisms. I mean, drinking the real christians of of of Churchill quite easy to to sum up and I do so in the war in the West, I say these are the things that they now use against him. I didn't do enough to avert the Bengal famine in 1943, for instance, that has been shot down by numerous historians, including Indian historians in the middle of the war, in the middle of a world war. Churchill did what he could to get grain supplies diverted From Australia to into two Bengal. Uh, the famine was appalling. It was caused by a typhoon, It was not caused by Winston Churchill and the idea that some basically indian nationalist historians have pumped out in recent years and just anti Churchill figures that he actually wanted indians to die as he's just total calumny. And when people claim some people claim that, I mean, there was a few very ignorant scholars were nevertheless with some credentials who claimed that Churchill wanted the indian population to basically genocide. Id. And it's complete nonsense, not least by the fact that during the period, which in question indian population boomed. So that's one of the main ones. Another one is that he had some views that we now regard as racist. He definitely regarded races as being of different characters. Um and that there were superior races and the as it were the white european was a superior culture. Um uh he was born in victorian England, so he had some victorian attitudes. Um these are things in the negative side of the ledger and as with all history, you should have a negative and a positive side of the ledger, positive side of the ledger includes, he almost certainly did more than any one human being to save the world from nazism. So that should count as something. And one of the reasons I talked about Churchill in this regard is to stress that if you get I'm not trying to stop anyone doing history at all. I don't think the revisionism of recent years about Churchill or the founding fathers of America or anyone else is anything I want to stop. I find it interesting find interesting, not least because it's so sloppy on occasions, but I find it interesting, it's important and we should be able to see people in the round. But that includes recognizing the positive side of the ledger and if you can't recognize that side, you're doing something else, you're doing something else. It's not history, it's some form of politicking of a very particular kind and I think it's the same thing with the founding fathers, There are some people, for instance, 30 since the 90s who have pushed the sally hemings Thomas Jefferson story to show that Thomas Jefferson was some kind of brute as a result. Um you know, we see Jefferson's statue being removed from the council chamber of the city was sitting in last november by council members who said that thomas jefferson no longer represents our values. If you can't if you can't recognize greatness of thomas Jefferson and that he had flaws. I mean that's not a grown up debate </s><s>[INST] and weigh them and weigh them in the context of the time. But let me sort of throw a curveball a curveball at you then what about recognizing the positive and the negative of a fellow with nice facial hair called Karl Marx? [/INST] Sure, sure. I mean I I have a section in the war in the West as you know where I go for Karl Marx with some glee. So </s><s>[INST] he seems to have gotten you know, some popularity in in the west recently, not [/INST] just recently. Yeah, I mean he's he's had a resurgence recently. </s><s>[INST] Well [/INST] that's because that's because whenever whenever things are seem to go wrong, people reach for other options and when for instance it's very hard for people to accumulate capital, it's not obvious that they're going to become capitalists. And so one thing that happens is people say, let's look at the Marxism thing against you if that's a viable go er My argument would simply be Point me to one place that's worked. </s><s>[INST] Well the argument from the Marxist or the Marxian economists is that we've only really tried it once. The Soviets tried it. And then if there's a few people that kind of tried the soviet thing, well they basically it's an offshoot of this show soviet. Yes. They've tried [/INST] it, tried it in Venezuela. </s><s>[INST] Yes. Yes. Yes. [/INST] So but let's just let's just quickly say how did all these experiments go? </s><s>[INST] They did not. Well they failed in fascinating ways [/INST] they did but they failed. </s><s>[INST] Yes. [/INST] And we should stress so grossly failed. So grossly failed that they threw millions and millions of people into completely thwarted lives that were much shorter than they should have been. </s>

<s>[INST] vaccine. [/INST] Yeah, yeah. Covid. Okay. And here's my point earlier, is that what I see is that the political divide has, has like a whirlpool that's pulling everything into it. And in that whirlpool thinking is done with the primitive mind tribes. And so I get, you know, okay, obviously something like race, right? That makes sense. That, that also right now the topic of race for example, or gender. These things are in the world pool, but that at least is like, okay, that's something that the primitive mind would always get really worked up about. You know, it's it it taps into like our deepest kind of like primal selves. Um Covid, you know, make this covid in a way to like, you know, climate change. Like that should just be something that our rational brains are like, let's solve this complex problem. But the problem is that it's all gotten sucked into the red versus blue whirlpool and once that happens it's in the hands of the primitive minds and we're losing our ability to be wise together to make decisions uh where it's like, it's like it's like the big species, brain is like the big american brain is like drunk at the wheel right now and we're about to go into a future with more and more big technologies, uh scary things, but to make big right decisions and not, you know, we're getting dumber as a as a collective. And that that's part of this environmental problem. </s><s>[INST] So within the space of technologists and the space of scientists, we should allow the arrows. That's one of the saddest things to me about is like the scientists, like I I've seen arrogance, there's a lot of mechanisms that maintain the tribe. It's the arrogance. It's it's a it's how you build up the this mechanism that defends this wall defends against the arrows. It's arrogance, credential ism like um just ego really and then just it protects you from actually challenging your own ideas. This ideal of science that makes science beautiful in in a time of fear and in a time of division created by perhaps politicians that leverage the fear like you said, makes the whole system dumber the science system dumber. Uh the the the tech developers system dumber if they don't allow the challenging of ideas [/INST] and what's really bad is that like in a normal environment you're always gonna have echo chambers And so what's the opposite of an echo chamber? I created a term for it, I think we need it which is called an ideal lab and idealab right? It's like people treat it like people act like scientists, even if they're not doing science, they just treat their ideas like science experiments and they toss them out there and everyone disagrees and disagreement is like the game, everyone likes to disagree, you know, in a certain text thread where everyone is just, you know, saying, you know, it's almost like someone throws something out and just as an impulse for the rest of the group to say, I think you're being like overly general there, I think like aren't you kind of being, I think that's like your bias showing and it's like no one's getting offended because it's like we're all just missing. We all of course respect each other. Obviously we're just, we're just, you know, trashing each other's ideas and that the whole group becomes smarter, you're always gonna have idea labs and echo chambers right in different communities and most of us participate in both of them, you know, and you know maybe in your marriage is a great idea, lab, you love to disagree with your your spouse and maybe in but this group of friends or your family at home, you know, you know in front of that sister, you do not bring up politics because she's now enforced. When that happens. Her bullying is forcing the whole room to be an echo chamber to appease her. Now what's what scares me is that usually have these things existing kind of in bubbles and usually there's like an h have their natural defenses against each other. So and an echo chamber person stays in their echo chamber. They don't like they will cut you out, they don't like they don't like to be friends with people who disagree with them. You notice that they will cut you out, they cut out their parents if they voted for trump or whatever, right? So there are, that's how they do it. They will say I'm going to stay inside of an echo chamber safely. So my ideas which I identify with um because my primitive mind is doing the thinking are not going to ever have to get challenge because it feels so scary and awful for that to happen. But if they leave and they go into an ideal environment, they're gonna, people are gonna say what? No, they're gonna disagree and they're gonna say, and then the person is going to try to bully they're they're gonna say that's really offensive and people will say no, it's not and they're gonna, they're gonna immediately say these people are assholes, right? So the echo chamber person, uh it doesn't have much power once they leave the echo chamber. Likewise the idealab person and they have this great environment, but they go into an echo chamber where everyone else's and they do that, they will get kicked out of the group, they will get branded as something, you know, a deny or a racist uh you know, uh right wing or a radical, you know, these these these nasty words. Um the thing that I don't like right now is that the echo chambers have found ways to forcefully expand into places that normally are have a pretty good immune system against echo chambers, like universities, like science journals, places where usually it's like there's a strong idealab culture there veritas, you know, you know, that that's that's an idealab slogan. Um you have is that these people have found a way to a lot of people have found a way to actually go out of their thing and keep their echo chamber by making sure that everyone is scared because they can punish anyone whether you're in their community or not. </s>

<s>[INST] because there's a huge even google and so on youtube transcription. Yeah, it's unclear but some of it is also integrating into a bigger system that so the user interface, how is deployed and all that kind of stuff. Maybe running it as an independent thing is much easier, like an order of magnitude easier than deploying to a large integrated system like Youtube transcription or um anything like meetings like zoom has trans transcription that's kind of crappy. But creating interface where detects the different individual speakers, it's able to um display it in compelling ways running in real time, all that kind of stuff. Maybe that's difficult. But that's the only explanation I have because like um I'm currently paying quite a bit for human transcription, human captions annotation and like it seems like uh there's a huge incentive to automate that it's very confusing [/INST] and I think, I mean I don't know if you look at some of the whisper transcripts but they're quite good. They're good, </s><s>[INST] especially in tricky cases. I've seen uh whispers performance on like super tricky cases and it does incredibly well. So I don't know, a podcast is pretty simple. It's like high quality audio and you're speaking usually pretty clearly. And so I don't know it uh I don't know what open they eyes plans are either, [/INST] but there's always like fun, fun projects basically. And stable diffusion also is opening up a huge amount of experimentation, I would say in the visual realm and generating images and videos and movies. And so that's going to be pretty crazy. That's going to, that's going to almost certainly work. And it's going to be really interesting when the cost of content creation is going to fall to zero. Used to need a painter for a few months to paint a thing and now it's going to be speak to your phone to get your video. </s><s>[INST] Hollywood will start using that to generate scenes Um which completely opens up. Yeah. So you can make a like a movie like Avatar eventually for under $1 million, much [/INST] less. Maybe just by talking to your phone? I mean I know it sounds kind of </s><s>[INST] crazy and then there'd be some voting mechanism like how do you have, like, would there be a show on netflix that's generated completely automated lee. So [/INST] potentially. Yeah. And what does it look like? Also when you can generate on demand and it's uh and there's infinity of </s><s>[INST] it. Yeah. Oh man all the synthetic content. I mean it's humbling because we treat ourselves as special for being able to generate art and ideas and all that kind of stuff if that can be done in an automated way by Ai. Yeah. [/INST] I think it's fascinating to me how these uh the predictions of Ai and what is going to look like and it's going to be capable of are completely inverted and wrong and sci fi of 50s and 60s was just like totally not right. They imagined a i is like super calculating here improvers and we're getting things that can talk to you about emotions. They can do art. It's just like weird. </s><s>[INST] Are you excited about that future? Just a eyes like hybrid systems, heterogeneous systems of humans. And Ai is talking about emotions. Netflix and Children ai system. That word. The netflix thing you watch is also generated by ai [/INST] I think it's it's going to be interesting for sure. And I think I'm cautiously optimistic but it's not it's not obvious. Well </s><s>[INST] the sad thing is your brain and mine developed in a time where before twitter before the before the internet. So I wonder people that are born inside of it might have a different experience. Um Like I maybe you can will still resisted. Uh and the people born now will not. [/INST] Well I do feel like humans are extremely malleable. </s><s>[INST] Yeah [/INST] and </s><s>[INST] you're probably right. What is the meaning of life andre [/INST] we we talked about sort </s><s>[INST] of the universe having a conversation with us humans or with the systems we create to try to answer for the university for the creator of the universe to notice us, we're trying to create systems that are allowed enough. Just answer back. [/INST] I don't know if that's the meaning of life, that's like meaning of life for some people, the first level answer I would say is anyone can choose their own meaning of life because we are conscious entity and it's beautiful number one. But I do think that like a deeper meaning of life if someone is interested is along the lines of like what the hell is all this? And like why? And if you look at the into fundamental physics and quantum field theory and the standard model, they're like very complicated and There's this like you know, 19 free parameters, parameters of our universe and like what's going on with all this stuff and why is it here? And can I hack it, can I work with it? Is there a message for me? AM I supposed to create a message? And so I think there's some fundamental answers there but I think there's actually even like you can't actually really make dent in dose without more time. And so to me also there's a big question around just getting more time honestly. Yeah, that's kinda like what I think about quite a bit as well. </s><s>[INST] So kind of the ultimate or at least first way to sneak up to the why question is to try to escape. Uh, the system, the universe and then for that, you sort of backtrack and say, okay for that, that's gonna be take a very long time. So the white question boils down from an engineering perspective to how do we extend? [/INST] Yeah, I think that's the question, number one, practically speaking, because you can't, you're not going to calculate the answer to the deeper questions in time you have and </s>

<s>[INST] that's fake. And then I [/INST] realized um there's all these like weird bitterness and like aggro nous and art about authenticity. But I had </s><s>[INST] this kind of like weird [/INST] realization recently uh </s><s>[INST] where I started thinking [/INST] that like </s><s>[INST] art [/INST] is sort of a decentralized collective </s><s>[INST] thing. [/INST] Like um </s><s>[INST] like art is kind of a [/INST] conversation with all the artists that have ever lived before. You. </s><s>[INST] You know like [/INST] it's like you're you're really just sort of it's not like anyone's reinventing the wheel here. Like you're kind of just taking </s><s>[INST] you know thousands of years of art [/INST] and um like running </s><s>[INST] it through your own [/INST] little algorithm and then like </s><s>[INST] making your like [/INST] your interpretation of it. You're just joining the </s><s>[INST] conversation with all the other [/INST] artists that came before. It's such a beautiful </s><s>[INST] way to look at it. Like [/INST] and it's like it's like I feel like everyone's always like there's </s><s>[INST] always copyright and I. P. [/INST] And this and that and </s><s>[INST] or authenticity [/INST] And it's just like like </s><s>[INST] I think we need [/INST] to stop seeing this as this like egotistical thing of like oh the </s><s>[INST] creative genius. The lone creative genius [/INST] or this or that because </s><s>[INST] it's like I think art isn't shouldn't [/INST] be about that. I think art is something that's sort </s><s>[INST] of brings humanity together. And it's [/INST] also art is also kind of the collective memory of humans. It's like we </s><s>[INST] don't like we don't give [/INST] a funk about whatever ancient Egypt like how much grain got sent that day and sending the records and like you know like </s><s>[INST] who went where and [/INST] you know how many shields needed to be produced for this? Like we just </s><s>[INST] remember there are art [/INST] and it's like you know it's like in our day to day life there's all this stuff that </s><s>[INST] seems more important [/INST] than art um because it helps us function and survive. But when all this is gone, like the only thing that's </s><s>[INST] really going to be left [/INST] is the art. The technology will be obsolete. </s><s>[INST] Like the humans [/INST] will be dead. That </s><s>[INST] is true [/INST] compression of human </s><s>[INST] history is the art we've [/INST] generated across the different uh centuries of different different millennia. So when the aliens come when the aliens </s><s>[INST] come, they're going to find the hieroglyphics [/INST] in the pyramids. I mean art can be broadly defined. They might find like the </s><s>[INST] engineering marvels the bridges [/INST] the rockets the </s><s>[INST] I guess I sort of [/INST] classify the architecture is art too. I consider engineering </s><s>[INST] um [/INST] in those formats to be art for sure. It sucks that like digital </s><s>[INST] art is easier to delete. So [/INST] if there's an apocalypse in nuclear war that can disappear </s><s>[INST] and [/INST] the physical, there's something </s><s>[INST] still valuable about the [/INST] physical manifestation </s><s>[INST] of art. That's, that [/INST] sucks that like </s><s>[INST] music for example, has to [/INST] be played by somebody. </s><s>[INST] Yeah, [/INST] I mean I do think we should have a foundation type situation where we like, you </s><s>[INST] know how we have like seed banks [/INST] up in the north and stuff. Like we </s><s>[INST] should probably have like, like [/INST] a, like a solar powered </s><s>[INST] or geothermal [/INST] little bunker that like has all the </s><s>[INST] all human knowledge. You [/INST] mentioned Daniel, I can Spotify. Um </s><s>[INST] what [/INST] do you think about that as an </s><s>[INST] artist? What [/INST] Spotify is that empowering? </s><s>[INST] Like to me, Spotify user as a consumer is super [/INST] exciting. It makes it easy for me to access music from </s><s>[INST] all kinds of artists, [/INST] get to </s><s>[INST] explore all kinds of [/INST] music, make it super </s><s>[INST] easy to sort [/INST] of curate my own playlists and have fun with all that. It was so liberating to let go. </s><s>[INST] You know, I [/INST] used to collect, you know, albums and cds and so on. Like, like, like hoard </s><s>[INST] albums, like [/INST] they matter. But the reality, you can, </s><s>[INST] you know, that was [/INST] really liberating, I can let go of that and letting go of the albums, </s><s>[INST] you're kind of [/INST] collecting allows you to find </s><s>[INST] new music, [/INST] exploring new artists and </s><s>[INST] all that kind of stuff. [/INST] But I know from a perspective an artist that could be, like you mentioned, </s><s>[INST] competition could [/INST] be a kind of constraint </s><s>[INST] because [/INST] there's more and more and more </s><s>[INST] artists [/INST] on the platform. </s><s>[INST] I think it's better that there's more [/INST] artists. I mean, again, this might be propaganda because this is </s><s>[INST] all from a conversation with Daniel [/INST] Ek, so this could easily be </s><s>[INST] propaganda. Like we're all [/INST] a victim of </s><s>[INST] somebody's propaganda. [/INST] So let's just accept this. </s><s>[INST] But Daniel [/INST] Ek was telling me that, uh, you know, at the because I, you know, when, when I </s><s>[INST] met him, I like [/INST] I I came in all furious </s><s>[INST] about Spotify [/INST] and like I grilled him super hard. So I've got his, </s><s>[INST] his, um answers [/INST] here. But </s><s>[INST] he was [/INST] saying like at the, the sort of peak of the </s><s>[INST] cd industry, [/INST] There was like 20,000 artists making millions and millions of </s><s>[INST] dollars, like there was just like a very tiny [/INST] Kind of 1%. Um, and Spotify is kind of democratized the </s><s>[INST] industry. Um, because [/INST] now I think he said there's about a million artists making a good living from Spotify. And when I heard that I </s><s>[INST] was like, honestly, [/INST] I would rather make less money and have </s><s>[INST] just like a decent [/INST] living. Um, then and have </s><s>[INST] more artists be [/INST] able to have that even </s><s>[INST] though I like, I wish it could [/INST] include everyone. But yeah, that's really hard to argue with youtube </s><s>[INST] is the same [/INST] as youtube's mission. They want to basically have as many creators as possible to make a </s><s>[INST] living. Some kind of [/INST] living. Yeah. And that that's so hard to argue with </s><s>[INST] you. But I think there's [/INST] better ways to do it. My manager, I actually wish he </s><s>[INST] was here. I like I [/INST] would have brought him up. My manager is building an app that </s><s>[INST] um [/INST] can manage </s><s>[INST] you. So it'll [/INST] like help you organize your percentages and um get your publishing and dada dada dada. You can take out all the middlemen so you can have a much </s>

<s>[INST] Yeah, the fighting, the fighting thing is uh I think Youtube and twitter and Facebook are providing an incredible service to the world and they're all making mistakes of course, but they're doing an incredible job, you know, that I think deserves to be applauded and there's some degree of gravity, like it's a cool thing that that that's created and uh it shouldn't be monolithically fought against like facebook is evil or so on. Yeah, I might make mistakes, but I think it's an incredible service. I think it's world changing. Uh I mean I've I think facebook has done a lot of incredible, incredible things by bringing for example identity. You're like uh allowing people to be themselves, like their real selves in uh in the digital space by using their real name and the real picture. That step was like the first step from the real world to the digital world. That was a huge step that perhaps will define the 21st century and thus creating a digital identity. And there's a lot of interesting possibilities there that are positive of course, some things that are, are negative and having a good dialogue about that is great and I'm great that people like you are at the center of the teletext. That's awesome. [/INST] Right? I think it also um I also can understand, I think actually in the past, especially in the past couple of years um this rising awareness has been helpful like uses are also more and more recognizing that privacy is important to them. They should um maybe right, they should be owners after data. I think this definitely is very helpful and I think also this type of voice also and together with the regulatory framework and so on, also hope the companies to essentially put these type of issues at a higher priority and knowing that, but also it is their responsibility to to ensure that users are well protected. So I think it definitely the writing voice is super helpful and I think that actually really has brought the issue of data privacy and even the consideration of data ownership to the forefront to really much wider community. Um and I think more of this voice is needed, but I think it's just that we want to have a more constructive dialogue um to bring both sides together to figure out constructive solution. </s><s>[INST] So, another interesting space where security is really important is in the space of uh any kinds of transactions, but it could be also digital currency. So, can you maybe talk a little bit about Blockchain? Can you tell me what is a [/INST] Blockchain, I think the Blockchain where it itself is actually very overload this uh in general. Right? Yes. So in general and talk of Blockchain, we refer to this distributed luxury in a decentralized fashion. So essentially you have in the community often knows that come together and even though each one may not be trusted and as ELISA certain thresholds after set of nose, he behaves properly, then the system can essentially achieve certain properties. For example, in the distribute legislating you have you can maintain an immutable log and you can ensure that for example the transactions actually I agreed upon and then it's immutable and so on. </s><s>[INST] So first of all, what's the ledger? So it's a [/INST] it's like a database, it's like a data entry. </s><s>[INST] And so distributed ledger is something that's maintained across or is synchronized across multiple sources, multiple notes. And so where is this idea? How do you keep? So it's important ledger, a database to keep that uh to make sure. So what are the kinds of security vulnerabilities that you're trying to protect against in the context of the of a distributed ledger? [/INST] So in this case for example you don't want some malicious knows to be able to change the transaction logs and in certain cases called double spending, like you also cause you can also cause different views in different parts of the network and so on. </s><s>[INST] So the ledger has to represent if you're capturing like financial transactions as to represent the exact timing and exact occurrence and no duplicates all that kind of stuff. It has to be uh represent what actually happened. Okay so what are your thoughts on the security and privacy of digital currency? I can't tell you how many people write to me to interview various people in the digital currency space. There seems to be a lot of excitement there. Uh and it seems to be some of it, to me from an outsider's perspective, seems like dark magic. I don't know how secure. I think the foundation from my perspective of digital currencies that is you can't trust anyone. So you have to create a really secure system. So can you maybe speak about how what your thoughts in general about digital currency is and how you how it can possibly create financial transactions and financial stores of money in the digital space. [/INST] So you ask about security and privacy. And so so again as I mentioned earlier in security, we actually talked about two main properties um the integrity and confidentiality and so there's another one for availability, you want the system to be available. But here for the question you asked, let's just focus on integrity and confidentiality. So so for integrity of this distribution, essentially as we discussed, we want to ensure that the different notes uh right, so they have this consistent view usually is done through, we call a consensus critical uh that's the established this shared view on this legend and that you cannot go back and change it is immutable and so on. So so in this case then the security often refers to this integrity property. And essentially you're asking the question how much work, how how can you attack the system so that uh the an attacker can change the lock for example. </s>

<s>[INST] but it didn't really start out there because well, what what happened before [/INST] that? Well, we don't know. So this is where it gets into a lot of speculation. And </s><s>[INST] let's go, I mean, [/INST] let's go there. Okay, sure. So, you know, </s><s>[INST] nobody [/INST] Can prove what happened before T equals zero. And whether there are other universes out there. I like to say that these are sort of on the boundaries of science. They're not just ideas that we wake up at three in the morning to go to the bathroom and say, oh well let's think about what happened before the Big Bang. Or let there be a multiplicity of universes. In other words we have real testable physics that we can use to draw certain conclusions that are plausibility arguments based on what we know. Now, admittedly there are not really direct tests of these hypotheses. That's why I call them hypotheses. They're they're not really elevated to a theory because a theory in science is really something that has a lot of experimental or observational support behind it. So their their hypotheses but they're they're not unreasonable hypothesis based on what we know about general relativity and quantum physics. Okay. And they may have indirect tests in that if you adopt this hypothesis, then there might be a bunch of things you expect of the universe. And lo and behold that's what we measure. But we're not actually measuring anything at T less than zero or we're not actually measuring the presence of another universe in this multiverse. And yet there are these indirect ideas that stem fourth. So it's hard to prove uniqueness and it's hard to completely convince oneself that a certain hypothesis must be true. But you know, the more and more tests you have that it satisfies let's say there are 50 predictions it makes and 49 of them are in are things that you can measure. And then the 50th one is the one where you you want to measure the actual existence of that other universe or what happened before t equals zero and you can't do that. But but you've satisfied 49 of the other testable predictions. And so that's science right now. A conventional condensed matter physicist or someone who deals with real data in the laboratory might say oh you cosmologists. You know that's not really science because it's not directly testable. But I would say it's sort of testable but it's not completely testable. And so it's at the boundary. But it's not like we're coming up with these crazy ideas among them quantum fluctuations out of nothing. And then inflating into a universe with you might say, well you created a giant amount of energy but in fact this quantum fluctuation out of nothing. You know in a quantum way violates the conservation of energy. But you know who cares that was a classical law anyway. And then an inflating universe maintains whatever energy it had be at zero or some infinite sizable amount in a sense the stuff of the universe has a positive energy but there's a negative gravitational energy associated with him. It's like I drop an apple. I got kinetic energy energy of motion out of that. But I did work on it to bring it to that height. So by going down and gaining energy of motion positive 12345 units of kinetic energy? It's also gaining or losing depending on how you want to think of it negative 12345 units of potential energy. So the total energy remains the same and inflating universe can do that. Or other physicists say that energy isn't conserved in general relativity. That's another way out of creating a universe out of nothing, you know? But the point is that this is all based on reasonably well tested physics. And although these extrapolations seem kind of outrageous at first, they're not completely outrageous there within the realm of what we call science already. And maybe some young whippersnapper will be able to figure out a way to directly test what happened before to equal zero or to test for the presence of these other universes. But right now we don't have a way of doing </s><s>[INST] that. So, speaking of young whippersnappers, roger Penrose, Yeah. So he kind of has a, you know, idea that we there might be some information that travels from whatever the heck happened before the Big Bang is that isn't about it. So, do you think it's possible to detect something like actually experimentally be able to detect some I don't know what it is radiation. Some some sort [/INST] of the cosmic </s><s>[INST] microwave background radiation. [/INST] There may be ways of doing </s><s>[INST] that. But is it is it philosophically or practically possible to detect signs that this was before the Big Bang? Or is it or is it what you said? Which is like everything we observe will, as we currently understand, we'll have to be a creation of this particular observable universe. [/INST] Yeah, I mean, you know, if you it's very difficult to answer right now, because we don't have a single verified, fully, self consistent, experimentally tested quantum theory of gravity. And of course, the beginning of the universe is a large amount of stuff in a very small space. So you need both quantum mechanics and general relativity. Same thing if our universe re collapses and then bounces back to another Big Bang, you know, there's also ideas there that some of the information leaks through or survives. I don't know that we can answer that question right now because we don't have a quantum theory of gravity that most physicists believe in. And belief is perhaps the wrong word that most physicists trust, because the experimental evidence favors it, right? You know, there are various forms of string theory. There's quantum loop gravity. There are various ideas but which, if any, Will be the one that survives the test of time and more importantly, within that, the test of experiment and observation. So, my own feeling is probably these things don't survive. I don't think we've seen any evidence in the cosmic microwave background radiation of of information leaking through similarly. Um, the one way or one of the few ways in which we might test for the presence of other universes as if they were to collide with ours that would leave a pattern of temperatures signature in the cosmic microwave background radiation. Some astrophysicists claim to have found it, but in my opinion, it's not statistically significant to the level that would be necessary to have such an amazing claim. Right? You know, it's just a 5% chance that the microwave background had that distribution just by chance. 5% isn't very long odds if you're claiming that instead that you're that you're finding, you know, evidence from another universe. I mean, it's like if the large hadron collider people had claimed after gathering enough data to show the Higgs particle, when there was a 5% chance, it could be just a statistical fluctuation in their data. No, they required five sigma five standard deviations, which is roughly one chance in two million that this is a statistical fluctuation of no physical </s>

<s>[INST] So this says something interesting about the human mind that as a computer is designed to be general. So that means you can so it's not it's not easily made specialized Meaning if you're going to be good at one thing uh Miyamoto Musashi has, this quote is an ancient warrior famous for the book of five rings in the martial arts world. And the quote goes if you know the way broadly you will see it in everything meaning if you do one thing is going to generalize to everything. And that that's an interesting thing about the human mind. So that that's what the G factor reveals. Okay so what's the difference if you can elaborate a little bit further between I. Q. And G. Factor? Just because it's a source of confusion for people [/INST] and I. Q. Is is a score. People use the word I. Q. To mean intelligence. But I. Q. Has a more technical meeting for people who work in the field and I it's an I. Q. Score score on a test that estimates the G. Factor. Um And the G factor is what's common among all these tests of mental ability. So if you think about it's not a Venn diagram but um I guess you could make a Venn diagram out of it but the g factor would be really at the core what's what's common to everything and I what I. Q. Scores do is they allow a rank order of people on the score and this is what makes people uncomfortable. This is where there's a lot of controversy about whether I. Q. Tests are biased toward any one group or another and um a lot of the the answers to these questions are very clear but they also have a technical aspect of it that's not so easy to to explain. </s><s>[INST] Well we'll talk about the fascinating and the difficult things about all of this. But so by the way when you say rank order that means you get a number and that means one person. You can now compare like uh you can say that this other person is more intelligent than me. What [/INST] you can say is I. Q. Scores are interpreted really as percentiles. So that uh if you have an I. Q. Of 100 and 40 and somebody else has 70. The metric is such that you cannot say the person with an I. Q. Of 100 and 40 is twice as smart As a person with an i. q. of 70. That would require a ratio scale with an absolute zero. Now you may think you know people with zero intelligence but in fact there is no absolute zero on on an I. Q. Scale it's relative to other people. So relative to other people. Somebody with an I. Q. Score of 140 is in the upper less than 1%. Whereas somebody with an I. Q. Of 70 is two standard deviations below the mean that's that's a different percentile. So </s><s>[INST] it's similar to like in chess you have an elo rating that's designed to rank order people. Uh So you can't say it's twice one person if if your your ratings twice another person, I don't think you're twice as good at chess. It's not stable in that way. But because it's very difficult to do these kinds of comparisons. But uh so what can we say about the number itself [/INST] is </s><s>[INST] that stable across tests and so on. And know [/INST] there are a number of statistical properties of any tests are called psychometric properties. You have validity, you have reliability, reliability. There are many different kinds of reliability. They all essentially measure stability and I. Q. Tests are stable within an individual. There are some longitudinal studies where Children were measured at age 11 and again when they were 70 years old and the two I. Q. Scores are highly correlated with each other. This comes from a fascinating study from Scotland. Uh in the 19 thirties some researchers decided to get an I. Q. Test on every single child age 11 in the whole country and they did. And those records were discovered in an old store room at the University of Edinburgh by a friend of mine Ian Deary who found the records digitized them. And has done a lot of research on the people who are still alive today from that original study including brain imaging research by the way. Really it's a fascinating group of of people who are who are are studied. Um not to get ahead of the story. But one of the most interesting things they found Is a very strong relationship between I. Q. measured at age 11 and mortality. So that You know in the 70 years later they looked at the survival rates and they could get death records from everybody. And Scotland has universal health care for everybody. And it turned out if you divide people by their age 11 I. Q. Score into court tiles. And then look at how many people are alive 70 years later. The I know this is in the book that I have the graph in the book but there are essentially twice as many people alive and the highest I. Q. Quartile than in the lowest quartile through in men and women. Um </s><s>[INST] interesting. So [/INST] it makes a big difference now why this is the case is not so clear since everyone had access to health care. </s>

<s>[INST] no, no. I I think about it a lot. I mean that's a really good thing to meditate on. We can talk about maybe how while you bring that up, what if it is or not a heavy question, but I get uh I think about mortality a lot and uh for me it feels like a really good way to focus in on is what you're doing today, The people you have around you, the family you have, is it uh does it bring you joy? Does it bring you fulfillment? And basically uh for me of long ago, okay, try to be ready to die any day. So like today, uh you know, I, I kind of woke up, look if I was nervous about talking to um have I really admire your work. And the book is very good and super exciting topic. Uh but then you know, there's this also feeling like if this is the last conversation I have in my life, you know if I die today, let's be let's be uh the right like my glad today happened and it is and I am glad today uh happened. So that that's the way [/INST] that's so unique. I never got that answer from a single person. </s><s>[INST] The business of life. There's goals, their dreams, there's like planning [/INST] very few people make it happen. That's what I learned. And so a lot of these people </s><s>[INST] like you run out of time, [/INST] it's not so much amount of time, but I'd come back later and be well okay, why don't you do that? If that's what you would do if you're going to die a year from now, why don't you, why don't you make it real simple things spend more time with family. Like why why don't you do that? And no one had an answer. It turns out unless you usually, unless you have, you really do have a pressing end of life. People don't do their bucket list or try to change their career. And some people can't so we can't like for a lot of people they can't do anything about it and that's that's fine. But the ones who can take action for some reason never do. And that was one of the ways that mike's death at the time, his impending death really really affected me because you know for these sick people what I learned, he had a bucket list and he was able to do some of the bucket list. It was awesome. But he got sick pretty quickly. So if you do only have a year to live, it's ironic because you can't do, you can't do the things you wanted to do because you get too sick too fast. </s><s>[INST] What were the bucket list things for you that you realize like what am I doing with my life? [/INST] That was the major concept of him after he died. I didn't know like I I was just lost because when something that profound happens all the things I was doing, most of the things I was doing were just meaningless was so tough to, to find an answer for that. And that's when I settled on I'm going to devote the rest of my life to trying to find another earth and to find out too find that we're not alone. </s><s>[INST] What is that longing for connection with others? Um What's that about? What do you think? Why is that so full of meaning? [/INST] I don't know why. I mean I think it's how we're hardwired. Like one of my friends some time ago actually my dad died. He never heard someone say this before but he's like Sarah you know, why are we evolved to take death so harshly what kind of society would we be if we just didn't care people died. That would be a very different type of world. How would we as a species have got to where we are? So I think that is tied hand in hand with, Why do we, why do we see connection? It's just that what we were talking about before, that sub consciousness that we don't understand. </s><s>[INST] Yeah coupled, you know, the other side the flipside of the coin of connection and, and love is a fear of loss. It's like that was again, I don't know. It's so, it makes you appreciate the moment Is that the thing ends? [/INST] Yeah, it's definitely a hard won, the thing ends, but it's hard to not, you, you wouldn't want to limit like it's like my dog who I love so much. I'll start to cry like I can't think about the end. I know he'll age much faster than I will and some day will end, right? But it's too sad to think of. But should I not have got a dog, I should have not brought this sort of joy into my life because I know it won't be forever. It's </s><s>[INST] well, there's uh, there's a philosopher Ernest Becker who wrote a book, denial of death. I just, uh, and warm with the cores. And there's another book talks about terror management theory, Sheldon Solomon. I just talked to him a few weeks ago, uh, the brilliant philosopher psychologist, that their theory, whatever you make of it, is that the fear of death is at the core of everything, everything we do. So like you're that you think you don't think about the mortality of your dog, but you do and that's what makes the experience rich. Like there's this kind of like in the shadows looks the knowledge that this won't last forever and that makes every moment just special in some kind of weird way that the moments are special for us humans? Uh I mean sorry to use romantic terms like love, but what do you make, what did you learn about love from, from losing it from losing your husband? [/INST] Well, I learned to love the things I have more. I learned to love the people that I have more and to not let the little things bother me as much. </s>

<s>[INST] I think when I was I was also introduced at that time to the idea Jessica in judo, I think in jiu jitsu is the same, a lot of sports is probably the same, is there's ways to win to include kind of uh if I were to use a bad term stalling, which is like use strategy to slow down to destroy all the weapons your opponent has, and just to wait it out to sort of break your opponent by yeah, shutting down all their weapons, but not using any of your own. And now Travis was always going for he's of course really good at gripping and do that whole game, but he was going for the big throws and he was almost getting frustrated by a lot of the opponents. Remember uh Ola bischoff, I [/INST] think uh </s><s>[INST] in Germany from [/INST] Germany talented, </s><s>[INST] very incredible. I know he's very good at doing big throws and he's incredible judoka, but he was also incredible at just frustrating his opponents with like gripping and strategy and so on. And I just remember feeling the pain of this person like Travis who went through just, he broke like every part of his body, He went through so many injuries, just this person who dedicated his entire life to this moment In 2008, and in 2012 In 2016, just ever gave everything, you could see it on his face uh that you know, his weapons are being shut down and he's still pushing forward. He's still that both the frustration and the power. I mean, the kind of throw he does is his main one I think is the standing was called Sarah Nagy saying, keep on saying that was, that was the other thing is like the techniques he used was these big throws, that there's something to me about the ST again, fell in love with that throw, that's my, become my main throw. Standing scientology. That is like, [/INST] why, why do you favor the standing variation because of the amplitude, you get a more powerful uh </s><s>[INST] power. It's like, are you a fan of Koga? Yes, that's so that's when I Travis, So Koga and Travis opened up my uh [/INST] Travis uses the same gripping peasants for saying Adidas coca </s><s>[INST] all the same and the way he uses his hips and turns. And I remember like going to my judo club and other judo clubs and asked and they were all saying this is the wrong way to do it. The way Travis does is the wrong way to do it. And I remember, I've always been [/INST] Amazed by this, by the way. I don't mean to cut you off, but I could literally fill 20 hours of reproductions of people who will tell me that either my students or other great world champions, um, doing things wrong. Yeah. And I'm, I'm looking at them and I'm like, who would I rather trust here in their judgment Koga, who was one of the greatest throwers of all time? Or you, a recreational guy who couldn't throw my grandmother. Um, uh, I'm supposed to take your word over his. We'll say, don't listen to what people say. I'm going to give you a piece of advice here. Watch what the best people do. Okay, that's how you get superior athletic performance. I'm going to say that again. Don't listen to what people say, watch what they do, particularly under the stress of high level competition because that's when you see their real game, what they really do under pressure again, if you can emulate that you're going to be very successful. </s><s>[INST] I guess what I was frustrated with to your point is that the argument against Koga is what he has a very specific body type and he figured out something that worked for him. The statement is that might not be applicable to you or to the general public of judo players that want to succeed that by the way at the shallow level might be true. Might be true. The point is there might be a body of knowledge that's yet to be discovered and explored. That Koga opened up that I wanted to understand why his technique worked. It made no sense to me that with a single foot, like the way you turn the hip, the single foot that steps in. Why does that work? Because it was actually very difficult to make work for me as a white belt in the very beginning, it doesn't make sense. Like people just, they don't they don't get loaded up onto your hip anyway. For people don't watch Koga highlights, Watch Travis stevens highlights. But the details of the technique don't make sense. But one mastered that. It feels like there's something fundamental there that hasn't been explored yet. It's like Koga and Travis made me think that we don't know most of the body mechanics involved in dominance in judo. Like we just kind of found a few pockets that worked really well. The Ramada, there's these different throws of Soto Gari. I wonder if there's like totally cool new things that we haven't discovered and that's saying I gave a little peak because there's very few people that I'm aware of that do it the way Travis and Koga did, [/INST] may I ask you a question. Yes. Um the choice of standing Sana g um I I should say this for your, for your listeners? They're probably thinking, what the hell are these two guys talking about? Um uh Zanardi is one of the more high percentage throws in the olympic sport of Judah. Um probably uh tomorrow is probably number one and variations of saying argue would be In the top five for sure. Um the basic choice you have in modern competition is the more difficult standing saying, argue where you literally are up on your feet and you perform a shoulder throw that takes your opponent over from a full standing position. The most popular form of saying R. G. And modern competition by a landslide is not the standing version. It's a drop saying argue where you go down to your knees. Um This means you have a much easier time getting underneath your opponent's central gravity. The defining feature of any Sana G is getting underneath your opponent's central gravity and lifting them to say Oy, military means to lift and carry. Um Why did you choose the more difficult version? What was your motivation? You know, you're smart kit, You know right from the start that for every standing sand are G. There's 20 drops Zanardi's and modern competition. One is obviously more high percentage. One obviously works for a wider variety of body types. The number of people who are successful with standings. Nrg is dramatically lower. Um And it appears to be a move which is completely absent in the heavyweight divisions, uh, and really seen in the lightweight divisions. Why what was the motivation? Why did you willingly adopt the less high percentage over </s>

<s>[INST] dying. There does seem to be a fundamental difference between the, the genghis khan style. Like if I would feel differently if somebody like hand to hand with a knife, murdered my family versus cybersecurity where I stole all their data stole all their money stole everything they own falsified their identity. All that kind of stuff that, those are both traumatic events. But they do seem to be fundamentally different. But maybe that's actually very narrow style thinking because ultimately you have to think about what his life and what's happiness and it's like the summer I thinking, I'm not sure what's more painful if I take it to myself, like, I'm not sure what I would rather lift through being stabbed like to death or having my identity stolen all my money stolen or maybe reputation destroyed? Like with lies or something like that. That's very interesting to think about, if you think about quality of life and all those kinds of things, [/INST] it's finite, Right. one point and the other is kind of a long prolonged, almost torture. </s><s>[INST] Physical pain versus psychological pain. It's really interesting to, it's really interesting to think [/INST] about. And I think another key piece to this, which I don't have answers to, but there is an element of emotion and rage in the physical violence versus again, more of that clinical, you know, information warfare. And so, um, it's really easy to show a battlefield where there's death on both sides and bombs being dropped in buildings destroyed. Um, it fits very well into propaganda for everyone involved, right? Regardless of what side they're on. When you start talking about cyber warfare, how many times have we seen a big company get data hacked or information hacked than we all read the headlines. Maybe throw a tweet out and then move on with our day and don't remember it anymore. And so it's just very, very different. I think in the motion and response that invokes and people when they hear about it as well, </s><s>[INST] given the conflict, do you think people are fundamentally good. Are we all sort of like blank slates [/INST] that could be </s><s>[INST] evil given the [/INST] environment </s><s>[INST] or good given the environment? Or can we kind of is there some bass that we can rely on [/INST] that people </s><s>[INST] if left to their own devices will be good and, and trustworthy and honest, [/INST] I think you have to separate out intention from action. So if you talk to some of the most heinous criminals in the world, they've rationalized their actions and so you've got to ask yourself what is the level of which they understood what they were doing, that they intended to be malicious, nefarious, um, and kind of do bad things versus the actual actions themselves. And so, um, even as a society, you know, if, let's say for example, you were to walk down the street and you were to murder somebody on the sidewalk completely unprovoked. We would look at you and say you are a, you know, a sociopath, You have everything wrong with you and that is somebody that we do not want in our society and therefore we will levy, you know, the extent of rule of law against you in order to protect society from you. Right? And you become that monster, that beast, if in the same situation, somebody walks in your house with a knife and you murder them or you kill them. Now we put you up on a pedestal as a hero and there are two very, very different responses. They may be just as morbid. They may be just as violent. Uh, in terms of the actual actions, the intentions, the way it's perceived is very, very different. And so I think that as a world we like a very black and white clean cut, you know, good, bad. I think the world is a lot more messy than that. And I think that a lot of times when you look at intention, it's hard for us to tell what somebody's intention is. But I've looked at a lot of, you know, people who are both considered very good and also a lot of people are considered very bad and they sound very similar in terms of the motivations for their actions. And so I think that it's, it's just a really hard problem that probably doesn't have a perfect solution or an answer. </s><s>[INST] You give much value to the intention. Or do you think it is better to look at the results of the behavior? Uh, as opposed to the, the underlying ideology, the underlying intention of the behavior. [/INST] I think that intention gets at the question of like, are people inherently good or bad, right, is I think that they're generally inherently good and the intention is driving towards the thing that they believe is the best outcome or the best thing for them to pursue the action I think is where we spend most of our time focused on. Um, and frankly we actually maybe a better society or kind of kinder as a humanity to each other, if we spent more time looking at intention rather than action. Um but again, you know, if somebody walks down the street and murder somebody, uh it's really hard to have a conversation and it may be inappropriate for us to have a conversation about intention versus any level of of action. Right? </s><s>[INST] So, you're one of the prominent, I would say, even the faces of the world of Cryptocurrency, Bitcoin, that whole entire world, let's dive straight in and ask, uh do you think do you consider yourself a Bitcoin maximalist? [/INST] I think that the way I really look at it is I work backwards off of what is the maximal ism that I believe in, and the world that I believe in is kind of an automated world that is run on these open, decentralized protocols. Uh and what we ultimately do is we return sovereignty and individualism and kind of personal responsibility and liberty. Uh two people over institutions and what we end up doing is we end up kind of taking what has historically been a very analog or physical world economy, and geographic rule. And we then kind of put it in the cloud, and this digital economy becomes the prevalent way that we all uh conduct commerce, communicate, et cetera. And so it's less about any one single technology to me, right? I think that it's pretty stupid for people to um almost in a way uh put an inanimate object up for uh you know, some sort of obsession Uh to me it's much more about the ideals, the ethos. Uh and the most rational and likely path to that kind of end result or that world that I think, you know, is at this point, just a foregone conclusion. So </s>

<s>[INST] What's the difference between a state of the art human such as you and I Well I don't know if you qualify state of their humans but a state of the art human and the state of the art robot. [/INST] That's a very interesting question. Robot capability is it's kind of uh I think it's a very tricky thing to to understand because there are some things that are difficult that we wouldn't think are difficult and some things that are easy that we wouldn't think are easy. Um And there's also a really big gap between capabilities of robots in terms of hardware and their physical capability and capabilities of robots in terms of what they can do autonomously. There is a little video that I think robotics researchers really like to show a special robotics learning researchers like myself from 2000 and four from stanford uh which demonstrates a prototype robot called the P. R. One. And the P. R. One was a robot that was designed as a home assistance robot. And there's this beautiful video showing the pr one tidying up a living room putting away toys and at the end of bringing a beer to the person sitting on the couch which looks really amazing. And then the punch line is that this robot is entirely controlled by a person. So you can so that in some ways the gap between the state of the art human state of the art robot if the robot is a human brain is actually not that large. Now obviously like human bodies are sophisticated and very robust and resilient in many ways. But on the whole, if we're willing to like spend a bit of money and do a bit of engineering, we can kind of close the hard work out almost. But the intelligence gap that one is very wide. </s><s>[INST] And when you say hardware, you you're referring to the physical sort of the actuators, the actual body, the robot is supposed to the hardware on which the cognition, The nervous, the hardware of the nervous [/INST] system. Exactly. I'm referring to the body rather than the mind. So that means that the kind of the work is cut out for us. Like while we can still make the body better, we kind of know that the big bottleneck right now is really the mind </s><s>[INST] and how big is that gap? How big is the, how big is the difference in your in your sense of ability to learn, ability to reason, ability to perceive the world between humans and our best robots. [/INST] The gap is very large and the gap becomes larger. The more unexpected events can happen in the world. So essentially the spectrum along which you can measure the size of that gap is the spectrum of how open the world is. If you control everything in the world very tightly if you put the robot in like a factory and you tell it where everything is and you rigidly program its motion, then it can do things. You know, one might even say in a superhuman way. It can move faster and stronger. It can lift up a car and things like that. But as soon as anything starts to vary in the environment now it'll trip up. And if many, many things vary like they would like in your kitchen for example, then things are pretty much like wide open. </s><s>[INST] No, again, we're gonna stick a bit on the philosophical questions. But how much on the human side of the cognitive abilities in your sense is nature versus nurture? So how much of it is a product of evolution? And how much of it's something we learn from sort of scratch well from the day we're born, [/INST] I'm going to read into your question is asking about the implications of this for a I because I'm not a biologist, I can't really speak authoritatively. </s><s>[INST] So until then Garnett, if if it's so if it's all about learning, then there's more hope for a I. [/INST] So the way that I look at this is that </s><s>[INST] um, [/INST] you know, well, first of course biology is very messy and it's if you ask the question, how does a person do something or has a person's mind do something? Um you can come up with a with a bunch of hypotheses and often times you can find support for many different, often conflicting hypotheses. </s><s>[INST] Um, [/INST] one way that we can approach the question of of of what the implication of this for a a ir is. We can think about what sufficient. So, you know, maybe a person is from birth. Very, very good at some things like for example, recognizing faces, there's a very strong evolutionary pressure to do that. If you can recognize your mother's face than you're more likely to survive and therefore people are good at this. But we can also ask like what's what's the minimum sufficient thing? And one of the ways that we can study the minimal sufficient thing is we could, for example, see what people do in unusual situations, if you present them with things that evolution could have prepared them for, you know, our our daily lives actually do this to us all the time. We we didn't evolve to deal with, you know, automobiles in spaceflight and whatever. Um, so there are all these situations that we can find ourselves in. Uh and we do very well there, like I can give you a joystick to control a robotic arm, which you've never used before. And you might be pretty bad for the first couple of seconds. But if I tell you like, your life depends on using this robotic arm to like open this door. You'll probably manage it even though you've never seen this device before. You've never used the choice to control us and you kind of muddle through it and that's not uh your evolved natural ability. That's your flexibility, your your adaptability and that's exactly what our current robotic systems really kind of fall flat. </s><s>[INST] But I wonder how much general, almost what we think of as common sense, pre trained models underneath all that. So that ability to adapt to a joystick, okay, It requires you to have a kind of, you know, I'm human, so it's hard for me to introspect all the knowledge I have about the world, but it seems like there might be an iceberg underneath of the amount of knowledge we actually bring to the table. That's kind of the open question. [/INST] I think there's absolutely an iceberg of knowledge that we bring to the table, but I think it's very likely that iceberg of knowledge is actually built up over our lifetimes because we have, you know, we have a lot of prior experience to drawn and it kind of makes sense that the right way for us to, you know, to optimize our our efficiency, our evolutionary fitness and so on, is to utilize uh all that experience to build up the best iceberg we can get. Uh and that's actually one of, you know, Well, that sounds an awful lot like what machine learning actually does. I think that for modern machine learning is actually a really big challenge to take this unstructured massive experience and distill out something that looks like a common sense understanding of the world and perhaps part of that is it's not because something about machine learning itself is broken or hard, but because we've been a little too rigid in subscribing to a very supervised, very rigid notion of learning, you know, kind of the input output excess go to y sort of model. And maybe what we really need to uh to do is to view the world more as like a massive experience that is not necessarily providing any rigid supervision, but sort of providing many, many instances of things that could be. And then you take that and you distill it into some sort of common sense understanding. </s><s>[INST] I see well you're you're painting an optimistic, beautiful picture, especially from the robotics perspective because that means we just need to invest and build better learning algorithms. Figure out how we can get access to more and more data for those learning algorithms to extract signal from and then accumulate that iceberg of knowledge. It's a beautiful picture, the hopeful one. [/INST] I think it's potentially a little bit more than just that. Um and this is this is where we perhaps reach the limits of our current understanding. But one thing that I think that the research community hasn't really resolved in a satisfactory way is how much it matters where the experience comes from. Like, you know, do you just like download everything on the internet and cram it into essentially the 21st century analog of the giant language model and and see what happens or does it actually matter whether your machine physically experiences the world or in the sense that it actually attempts things, observes the outcome of its actions and and kind of augmented experience that way </s><s>[INST] that it chooses which parts of the world it gets to interact with and observe and learn from [/INST] right. It may be that the world is so complex that uh simply obtaining a large mass of sort of I. I. D. Samples of the world is a very difficult way to go. But uh if you are actually interacting with the world and essentially performing the sort of hard negative mining by attempting what you think might work observing the sometimes happy and sometimes sad outcomes of that and augmenting your understanding using that experience and you're just doing this continuously for many years, maybe that sort of data in some sense is actually much more favorable to obtaining a common sense understanding. One reason we might think that this is true is that um you know, the what we associate with common sense or lack of common sense is often characterized by the ability to reason about kind of counterfactual questions. Like you know, if I were to uh here, I'm this bottle of water sitting on the table, everything is fine for knocking over uh which I'm not going to do, but if I were to do that well what would happen uh and I know that nothing good will happen from that, but if I have a bad understanding of the world, I might think that that's a good way for me to like you know gain more utility. Uh If I actually go about my daily life doing the things that my current understanding of the world suggests will give me high utility in some ways I'll get exactly the right uh supervision to tell me not to do those bad things and to keep doing the good things. </s><s>[INST] So there's a spectrum between my I. D. Random walk through the space of data and then there's and what we humans do. Well I don't even know if we do it the optimal. But there might be beyond. So this open question that you raised where do you think systems intelligence systems that would be able to deal with this world fall? Can we do pretty well by reading all of Wikipedia sort of randomly sampling it? Like language models do or do we have to be exceptionally selective and intelligent about which aspects of the world we attract with. [/INST] So I think this is first an open scientific problem and I don't have like a clear answer but I can speculate a little bit and what I would speculate is that you don't need to be super super careful. I think it's less about like being careful to avoid the useless stuff and more about making sure that you hit on the really important stuff. So perhaps it's okay if you spend part of your day just you know guided by your curiosity visiting interesting regions of the of your state space. But it's important for you to you know, every once in a while, make sure that you really try out the uh solutions that your current model of the world suggests might be effective and observe whether those solutions are working as you expect or not. And perhaps uh some of that is really essential to have kind of a perpetual improvement loop. Like this perpetual improvement loop is really like that. That's really the key, the key that's going to potentially distinguish the best current methods from the best methods of tomorrow in a sense. How </s><s>[INST] important do you think is exploration are totally out of the box thinking, exploration in this space is jump two totally different domains. So you kind of mentioned, there's an optimization problem, you kind of kind of explore the specifics of a particular strategy, whatever the thing you're trying to solve, how important is it to explore totally outside of the strategies have been working for you so far. What's your intuition there? [/INST] Yeah, I think it's a very problem dependent kind of question and I think that that's actually, you know in some ways that question gets at um one of the big differences between uh sort of the classic formulation of a reinforcement learning problem and uh some of the sort of more open ended reformulation of that problem that has been explored in recent years. So classically reinforcement learning is framed as a problem of maximizing utility, like any kind of rational ai agent and then anything you do is in service to maximizing the utility. But a very interesting kind of way to look at. Um I'm not necessarily saying this is the best way to look at about an interesting alternative way to look at these problems is um as something where you first get to explore the world however you please and then afterwards you will be tasked with doing something and that might suggest a somewhat different solutions. So if you don't know what you're going to be tasked with doing and you just want to prepare yourself optimally for whatever you're uncertain future holds maybe then you will choose to attain some sort of coverage build up sort of an arsenal of cognitive tools if you will such that later on when someone tells you now your job is to fetch the coffee for me, you'll be well prepared to undertake that task </s><s>[INST] and that you see that as the modern formulation of the reinforcement learning problem as a kind of the more multi task, the general intelligence kind of formulation. [/INST] I think that's one possible vision of where things might be headed. I don't think that's by any means the mainstream or standard way of doing things and it's not like if I had to </s><s>[INST] but I like it. It's a it's a beautiful vision. So maybe you actually take a step back what is the goal of robotics? What's the general problem in robotics we're trying to solve? You actually kind of painted two pictures here. One of the narrow one of the general what in your view is the big problem with robotics? Again, ridiculously philosophic Ohio for questions. [/INST] I think that um, you know, maybe there are two ways I can answer this question. One is there's a very pragmatic problem which is like what would make robots well, sort of maximize the usefulness of robots. And they're, the answer might be something like a system where uh, a system that can perform whatever task a human user sets for it, you know, within the physical constraints. Of course, if you tell it to teleport to another planet but probably can't do that. But if you ask it to do something that's within its physical capability, then potentially with a little bit of additional training or a little bit of additional trial and error, it ought to be able to figure it out in much the same way as like a human telly operator ought to figure out how to how to drive the robot to do that. That's kind of a very pragmatic view of what it would take to kind of solve the robotics problem if you will. But I think that there is a second answer and that answer that, that the answer is a lot closer to why I want to work on robotics, which is that I think it's it's it's less about what it would take to do a really good job in the world of robotics but more the other way around what robotics can bring to the table to help us understand artificial intelligence. </s><s>[INST] So your dream fundamentalist to understand intelligence? [/INST] Yes. I think that's the dream for many people who actually work in this space. I think that there's there's something very pragmatic and very useful about studying robotics. But I do think that a lot of people that go into this field actually, you know, the things that they draw inspiration from are the potential for robots to help us learn about intelligence and about ourselves. So </s><s>[INST] that's fascinating that robotics is basically the space by which you can get closer to understanding the fundamentals of artificial intelligence. So what is it about robotics that's different from some of the other approaches? So if you look at some of the early breakthroughs in deep learning when the computer vision space and the natural language processing there was really nice clean benchmarks that a lot of people competed on and their body came off a lot of brilliant ideas. What's the fundamental difference between computer vision purely defined an Image net and kind of the bigger robotics problem. [/INST] So there are a couple of things. One is that with robotics um you kind of have um you kinda have to take away many of the crutches. So you have to deal with with both the the particular problems of perception control and so on. But you also have to deal with the integration of those things and you know classically we've always thought of the integration as kind of a separate problem. So a classic kind of modular engineering approaches that we solve the individual sub problems then wire them together and then the whole thing works. Um And one of the things that we've been seeing over the last couple of decades is that well maybe studying the thing as a whole might lead to just like very different solutions that we if we were to study the parts and wire them together. So the integrated nature of robotics research helps us see you know the different perspectives on the problem. Another part of the answer is that with robotics um It casts a certain paradox into very clever relief so that this is sometimes referred to as a more of a paradox. The idea that in artificial intelligence things that are very hard for people can be very easy for machines and vice versa. Things that are very easy for people can be very hard for machines. So you know uh integral and differential calculus is pretty difficult to learn for people but if you program a computer do it it can derive derivative and integral for you all day long without any trouble. Whereas some things like you know drinking from a cup of water very easy for a person to do very hard for a robot to deal with. And sometimes when we see such blatant discrepancies that give us a really strong hint that we're missing something important. So if we really try to zero in on those discrepancies we might find that little bit that were missing and it's not that we need to make machines better or worse at math and better at drinking water, but just that by studying those discrepancies we might find some new insight </s><s>[INST] so that that could be that could be in any space doesn't have to be robotics. But you're saying um yeah it's kind of interesting that robotics seems to have a lot of those discrepancies. So the hans more of a paradox is probably referring to the space, of the physical interaction, like you said, object manipulation, walking all the kind of stuff we do in the physical world. Yeah. Uh that how do you make sense if you were to try to disentangle on the the more box paradox like why is there such a gap in our intuition about it, Why do you think manipulating objects is so hard from everything you've learned from applying reinforcement learning in this space? [/INST] Yeah I think that one reason is maybe that for many of the for many of the other problems that we've studied in ai and computer science and so on. Okay, the notion of input output and supervision is much much cleaner. So computer vision for example, deals with very complex inputs, but it's comparatively a bit easier, at least up to some level of abstraction to cast it as a very tightly supervised problem. It's comparatively much much harder to cast robotic manipulation as a very tightly supervised problem. You can do it, it just doesn't seem to work all that well. So you could say that while maybe we get a labeled dataset where we know exactly which motor commands to send and then we train on that. But for various reasons, that's not actually like such a great solution. And it also doesn't seem to be even remotely similar to how people and animals learn to do things because we're not told by like our parents here is how you fire your muscles in order to walk. We, you know, we do get some guidance but the really low level detailed stuff we figure out mostly on our own. </s><s>[INST] And that's what you mean by tightly coupled at every single little sub Action gets a supervised signal of whether it's a good one or not. [/INST] Right? So, so while in computer vision you could sort of imagine up to a level of abstraction that maybe, you know, somebody told you this is a car and this is a cat and this is a dog in motor control. It's very clear that that was not the case </s><s>[INST] if you look at sort of the sub spaces of robotics that, again, as you said, robotics integrates all of them together and we'll get to see how this beautiful mess into place. But so there's nevertheless still perception. So it's the computer vision problem, broadly speaking, understanding the environment. Then there's also maybe you can correct me on this kind of categorization of the space. Then there's prediction in trying to anticipate what things are going to do into the future in order for you to be able to act in that world. And then there's also this game theoretic aspect of how your actions will change the behavior of others. Oh in this kind of space what? And this is bigger than reinforcement learning. This is just broadly looking at the problem in robotics. What's the hardest problem here? Or is there or is what you said true that when you start to look at all of them together, that's a that's a whole another thing like you can't even say which one individually is harder because all of them together, you should only be looking at them all together. [/INST] I think when you look at them all together some things actually become easier and I think that's actually pretty important. So we had, you know back in 2014 we had some more basically our our first work on end and reinforced learning for robotic manipulation skills from vision which you know at the time was something that seemed a little inflammatory and controversial in the robotics world, but uh other than the the inflammatory controversial part of it, the point that we were actually trying to make a network is that for the particular case of combining perception and control, you could actually do better if you treat them together than if you try to separate them. And the way that we try to demonstrate this is we picked a fairly simple motor control task where a robot had to insert a little red trapezoid into a trapezoidal whole. Uh and we had our separated solution which involved first detecting the whole using opposed detector and then I actuated the arm to put it in and then our intense solution which just map pixels to the torques. Uh And one of the things we observed is that if you use the intense solution essentially the pressure on the perception part of the model is actually lower. Like it doesn't have to figure out exactly where the thing is in three D. Space. It just needs to figure out where it is. You know, distributing the errors in such a way that the horizontal difference matters more than the vertical difference because vertically just pushes it down all the way until it can't go any further and their perceptual errors are a lot less harmful. Whereas perpendicular to the direction of motion, perceptual errors are much more harmful. So the point is that if you combine these two things, you can trade off errors between the components optimally to best accomplish the task and the components actually be weaker while still leading to better overall performance as a </s><s>[INST] profound idea in in in the space of pegs and things like that is quite simple. It almost is attempting to overlook, but that's seems to be at least intuitively idea that should generalized basically all aspects of perception and control. Of course one strengthens the other. [/INST] Yeah. And and we, you know, people who have studied uh sort of perceptual heuristics and humans and animals find things like that all the time. So 11 very well known example, there's something called the gaze heuristic, which is a little trick that you can use to intercept the flying object. So if you want to catch a ball, for instance, you could try to localize it in three D. Space, estimate its velocity, estimate the effect of wind resistance, solve a complex system of differential equations in your head um or you can maintain a running speed so the object stays in the same position in your field of view. So if it dips a little bit, you speed up if it rise a little bit uh you slow down and if you follow this simple rule, you'll actually arrive at exactly the place where the object lands and you'll catch it and humans use it when they play baseball. Human pilots use it when they fly airplanes to figure out. If they're about to collide with somebody frogs uses to catch insects and so on and so on. So this is something that actually happens in nature. And I'm sure this is just one instance of it that we were able to identify just because all the scientists were able to identify because so prevalent, but they're probably many others. </s><s>[INST] Do you have a, just so you can zoom in as we talked about robotics, You have a canonical problem sort of a simple, clean, beautiful representative problem and robotics that you think about when you're thinking about some of these problems, we talked about robotic manipulation. To me that seems intuitively at least the robotics community has converged towards that as a space. That's the canonical problem. If you agree that maybe you zoom in in some particular aspect of that problem that you just like, like if we solve that problem perfectly, it'll unlock a major step in um, towards human level intelligence. [/INST] I don't think I have like a really great answer to that. And I think partly the reason they don't have a great answer kind of has to do with, it has to do with the fact that the difficulty is really in the flexibility and adaptability rather than doing a particular thing really, really well. So um, it's hard to just say like, oh, if you can, I don't know, like shuffle a deck of cards as fast as like a Vegas uh, casino dealer, then you'll, you'll be very proficient. It's really the ability to quickly figure out how to do some arbitrary new thing well enough to like, you know, to move on to the next arbitrary thing. </s><s>[INST] But the source of newness and uncertainty, have you found uh problems in which it's easy to uh generate new new Nous Nous Nous is Yeah, new types of newness. [/INST] Yeah, so a few years ago, so if you had asked me this question around like 2016, maybe, I would have probably said that robotic grasping is a really great example of that because it's a task with great real world utility, like you will get a lot of money if you can do it well. </s><s>[INST] What is robotic grasping, [/INST] picking up any object </s><s>[INST] with a robotic hand. [/INST] Exactly. So you'll get a lot of money if you do it well because lots of people want to run warehouses with robots and it's highly nontrivial because uh very different objects uh will require very different grasping strategies. But actually since then people have gotten really good at building systems to solve this problem uh to the point where I'm not actually sure how much more progress we can make with that as like the main guiding thing. But it's kind of interesting to see the kind of methods that have actually worked well in that space because robotic grasping classically used to be regarded very much as um kind of almost like a geometry problem. So uh people who have uh studied the history of computer vision will find this very familiar, that kind of in the same way that in the early days of computer vision, people thought of it very much as like an inverse graphics thing in robotic grasping, people thought of it as an inverse physics problem. Essentially, you look at what's in front of you figure out the shapes. Then use your best estimate of the laws of physics to figure out where to put your fingers on. You pick up the thing. Um and it turns out that what works really well for robotic grasping instance, created in many different recent works, including our own but also wants from many other labs, is to use learning methods with some combination of either exhaustive simulation or like actual real world trial and error. And turns out that those things actually work really well. And then you don't have to worry about solving geometry, problems of physics problems. </s><s>[INST] So what are just by the way on the grasping? What are the difficulties that have been worked on? So one is like the materials of things, maybe inclusions and the perception side. Why why is this such a difficult why is picking stuff up? Such a difficult problem? [/INST] Yeah, it's a difficult problem because the number of things that you might have to deal with or the variety of things you have to deal with is extremely large. And um oftentimes things that work for one class of objects won't work for other classes of objects. So if you if you get really good at picking up boxes and now you have to pick up plastic bags. Um you know, you just need to employ a very different strategy and there are many properties of objects that are more than just their geometry that has to do with, you know, the bits that are easier to pick up the bits that are hard to pick up. The bits that are more flexible. The bits that will cause the thing to pivot and bend and drop out of your hand versus the bits that result in a nice secure grasp, things that are flexible things that if you pick them up the wrong way they'll fall upside down and the contents will spill out. So there's all these little details that come up. But the task is still kind of can be characterized as one task. There's a very clear notion of you did it or you didn't do it. </s><s>[INST] So in terms of spilling things, they're creeps in this notion that starts to sound and feel like common sense reasoning. Do you think solving the general problem of robotics requires common sense reasoning requires general intelligence. This kind of human level capability of, you know, like you said, be robust and deal with uncertainty but also be able to sort of reason and assimilate different pieces of knowledge that you have. Um Yeah, what are you, what are your thoughts on the needs of common sense reasoning in the space of the general robotics problem. [/INST] So I'm gonna slightly dodge that question. And so that I think I think maybe actually it's the other way around is that studying robotics can help us understand how to put common sense into our ai systems. one way to think about common sense is that and why our current systems might lack common sense is that common sense is a property is an emergent property of um actually having to interact with a particular world, a particular universe and get things done in that universe. So you might think that for instance, like an image captioning system uh maybe it looks at pictures of the world and it types out english sentences. So it kind of, it kind of deals with our world and then you can easily construct situations where image captioning systems do things that defy common sense. Like give it a picture of a person wearing fur coat and we'll say it's a teddy bear. But I think what's really happening in those settings is that the system doesn't actually live in our world. It lives in its own world. That consists of pixels in english sentences and doesn't actually consist of like you know having to put on a fur coat in the winter so you don't get cold. So perhaps the reason for the disconnect is that the systems that we have now simply inhabit a different universe And if we build ai systems that are forced to deal with all of the messiness and complexity of our universe, maybe they will have to acquire common sense to essentially maximize their utility. Whereas the systems we're building now don't have to do that. They can take some shortcut. </s><s>[INST] That's fascinating. You've a couple times already sort of reframed the role of robotics and this whole thing. And for for some reason I don't know if I'm, my way of thinking is common but I thought like we need to understand and solve intelligence in order to solve robotics and you're kind of framing it as no robotics is one of the best ways to just study artificial intelligence and build sort of like robotics is like the right space in which you get to explore some of the fundamental learning mechanisms, fundamental sort of multi modal, multi task aggregation of knowledge, mechanisms that are required for general intelligence. That's really interesting way to think about it. But let me ask about learning can the general sort of robotics the epitome of the robotics problem be solved purely through learning perhaps end to end learning sort of learning from scratch as opposed to injecting human expertise and rules and heuristics and so on. [/INST] I think that in terms of the spirit of the question I would say yes. I mean I think that though in some ways it may be like an overly sharp dichotomy like you know, I I think that in some ways when we build algorithms we, you know, at some point a person does something like a person, a person turned on the computer, A person uh you know, implemented tensorflow. Um but yeah, I think that in terms of in terms of the point that you're getting and I do think the answer is yes. I think that I think that we can solve many problems that have previously required meticulous manual engineering through automated optimization techniques. And actually one thing I will say on this topic is, I don't think this is actually a very radical or very new idea. I think people, I have been thinking about automated optimization techniques as a way to do control for a very, very long time. And in some ways what's changed is really more the name. So, you know, today we would say that, Oh my robot does machine learning, it does reinforcement learning. Maybe in the 1960s you'd say, Oh my robot is doing optimal control. And maybe the difference between typing out a system of differential equations and doing feedback, linear ization versus training and neural net, maybe it's not such a large difference, it's just uh you know, pushing the optimization deeper and deeper into the thing. </s><s>[INST] Well, it's interesting you think that would, but with especially with deep learning that the accumulation of sort of experiences and data form to form deep representations starts to feel like knowledge is supposed to optimal control. This feels like there's an accumulation of knowledge to the learning process. [/INST] Yes. Yeah. So I think that is a good point that one big difference between learning based systems and classic optimal control systems is that learning based systems in principle should get better and better the more they do something. And I do think that that's actually a very very powerful difference. </s><s>[INST] So if we look back at the world of expert systems, the symbolic ai and so on. Of using logic to accumulate expertise, human expertise, human encoded expertise, do you think that will have a role at some point? That the you know deep learning machine learning reinforcement learning has been has shown incredible results and breakthroughs and just uh inspired thousands maybe millions of researchers. But you know there's this uh less popular now but it used to be a popular idea of symbolic ai do you think that will have a role? [/INST] I think in some ways the kind of the the descendants of uh symbolic I actually already have a role. So you know this is the highly biased history from my perspective you say that? Well initially we thought that rational decision making involves logical manipulation. So you have some model of the world expressed in terms in terms of logic. Uh You have some quarry like what action do I take in order to for X to be true and then you manipulate your logical symbolic representation to get an answer. What that turned into somewhere in the 1990s is uh well, instead of building kind of predicates and statements that have true or false values will build probabilistic systems where uh things have probabilities associated probabilities of being true and false and that turned into base nets. Uh and that provided sort of a boost to what we're really, you know, still essentially logical inference systems. Just probabilistic, logical inference systems. And then people said, well let's actually learn the individual probabilities inside these models. And then people said, well let's not even specify the nodes in the models. Let's just put a big neural net in there. But in many ways I see these as actually kind of descendants from the same idea. It's essentially extenuating rational decision making by means of some inference process and learning by means of an optimization process. So, so in a sense, I would say yes, that has a place and in many ways that places or you know, it already holds that place, </s><s>[INST] it's already in there. Yeah, it's just by different, it looks slightly different than it was before. [/INST] But but in some there are some things that that we can think about that make this a little bit more obvious. Like if I train a big neural net model to predict what will happen in response to my robot's actions and then I run probabilistic inference meaning I invert that model to figure out the actions that lead to some plausible outcome. Like to me that seems like a kind of logic, you have a model of the world, it just happens to be expressed by a neural net. And you are doing some inference procedures, some sort of manipulation on that model to figure out uh you know, the answer to a quarry that you have. </s><s>[INST] It's the interpret ability. It's the explain ability though, that seems to be lacking more so because the nice thing about expert systems is you can follow the reasoning of the system that to us, mere humans is somehow compelling. It would, it's just uh I don't know what to make of this fact that there's a human desire for intelligence systems to be able to convey in a poetic way to us why I made the decisions. It did like tell a convincing story and perhaps that's like a a silly human thing, Like we shouldn't expect that of intelligence systems, like we should be super happy that there's intelligence systems out there. But uh if I were to sort of psychoanalyze the researchers at the time, I would say expert systems connected to that part. That desire of ai researchers for systems to be explainable. I mean maybe on that topic, do you have a hope that inferences to solve learning based systems will be as explainable as the dream was with expert systems for example. [/INST] I think it's a very complicated question because I think that in some ways the question of explain ability is um, kind of very closely tied to the question of, of like performance. Like, you know, what, why do you want your system to explain itself well, so that it's so that when it screws up, you can kind of figure out why it did it. But in some ways that that's a much bigger problem actually, like your system might screw up and then it might screw up and how it explains itself or you might have some bugs somewhere so that it's not actually doing what was supposed to do. So, you know, maybe a good way to view that problem is really is a problem as a bigger problem of verification and validation um, of which explain abilities sort of one component. </s><s>[INST] I see, I I just see it differently. I see explain ability. You, you put it beautifully. I think you actually summarize the field of explain ability. But to me there's another aspect of explain ability which is like storytelling that has nothing to do with errors or with like the sort of, it doesn't, it uses errors as as uh, elements of its story as opposed to a fundamental need to be explainable when errors occur. It's just that for other intelligence systems to be in our world, we seem to want to tell each other stories and that's true in the political world is true in the academic world and that, you know, neural networks are less capable of doing that or perhaps they're equally capable of storytelling storytelling. Maybe it doesn't matter what the fundamentals of the system are. You just need to be a good storyteller. [/INST] Maybe one specific story I can tell you about in that space is actually about some work um That was done by uh by my former collaborator who's now a professor at M. I. T. Named Jacob Andreas. Um Jacob actually works in natural language processing but he had this idea to do a little bit of work in reinforcement learning and how um on how natural language can basically structure the internals of policies trained with RL and one of the things he did is he set up a model that attempts to perform some task that's defined by a reward function. But the model reads in a natural language instruction. So this is a pretty common thing to do an instruction following. So you tell it like you know go to the Red House and then I'm supposed to go to the Red House. But then one of the things that Jacob did is he treated that sentence not as a command from a person but as a representation of the internal kind of state of the of the of the mind of this policy essentially. So that when it was faced with a new task, what it would do is it would basically try to think of possible language descriptions, attempt to do them and see if they lead to the right outcome. So would kind of think out loud like you know, I'm faced with this new task, what am I gonna do? Let me go to the red house. No that didn't work. Let me go to the Blue room or something. Let me go to the green plant and once it got some reward would say oh go to the green plant, that's what's working. I'm gonna go to the green plant and then you could look at the string that it came up with and that was a description of how it thought it should solve the problem. So you could do, you could basically incorporate language as internal state and you can start getting some handle on these kinds of things </s><s>[INST] and then what I was kind of trying to get to is that also if you add to the reward function, the convincing nous of that story. So I have another award signal of like people who review that story how much they like it. I said that you know, initially that could be a hyper parameter sort of hard coded heuristic type of thing. But it's an interesting notion of the convincing nous of the story becoming part of the reward function. The objective function of the explain ability. It's in the world of sort of twitter and fake news. That might be a scary notion that the nature of truth may not be as important as the convincing nous of the how convinced you are in telling the story around the facts. Well, let me ask the basic question. You're one of the world class researchers and reinforcement learning deeper enforcement learning certainly in the robotic space. What is reinforcement learning? [/INST] I think that what reinforcement learning refers to today is really just the uh kind of the modern incarnation of learning based control. So classically reinforcement learning has a much more narrow definition which is that it's you know, literally learning from reinforcement like the thing does something and then it gets a reward or punishment. But really I think the way the term is used today is it's used to reform more broadly to learning based control. So some kind of system that's supposed to be controlling something and it uses data to get better </s><s>[INST] and what does control mean? So his action is the fundamental element. [/INST] It means making rational decisions now and rational decisions or decisions that maximize the measure of utility </s><s>[INST] and sequentially so many decisions. Time and time and time again now, like it's easy to see that kind of idea in the space of maybe games in the space of robotics. Do you see it bigger than that? Is it applicable? Like we're the limits of the applicability of reinforcement learning. [/INST] Yeah. So rational decision making is essentially the encapsulation of the ai problem viewed through a particular lens. So any problem that we would want a machine to do? Uh intelligent machine can likely be represented as a decision making problem. Classifying images is a decision making problem, although not a sequential one. Typically uh you know, controlling a chemical plant as a decision making problem, deciding what videos to recommend on youtube is a decision making problem. And one of the really appealing things about reinforcement learning is if it does encapsulate the range of all these decision making problems, perhaps working on reinforcement learning is uh you know, one of the ways to reach a very broad swath of ai problems. </s><s>[INST] But what what do you is the fundamental difference between reinforcement learning and maybe supervised machine learning. [/INST] So reinforcement learning can be viewed as a generalization of supervised machine learning. You can certainly cast supervised learning as a reinforcement learning problem. You can just say your loss function is the negative your reward. But you have stronger assumptions, you have the assumption that someone actually told you what the correct answer was uh that your data was idea and so on. So you could view reinforcement learning essentially relaxing some of those assumptions. Now, that's not always a very productive way to look at it because if you actually have a supervised learning problem, you'll probably solve it much more effectively by using supervised learning methods because it's easier. But you can view reinforced morning as a generalization of that. Know for sure, </s><s>[INST] but they're fundamentally different. That's a mathematical statement. That's absolutely correct, but it seems that reinforcement learning, the kind of tools will bring to the table today of today. So maybe down the line, everything will be a reinforcement learning problem just like you said, image classification should be mapped to a reinforcement learning problem. But today the tools and ideas, the way we think about them are are different sort of supervised learning has been used very effectively to solve basic narrow ai problems, reinforcement learning kind of represents the dream of a I it's very much so in the research space now into captivating the imagination of people of what we can do with intelligence systems, but it hasn't yet had as wide of an impact as a supervised learning approaches. So that sort of, my question comes to the more practical sense, like what do you see is the gap between the more general reinforcement learning and the very specific yes, it's a financial decision making with one sequence, one step in the sequence of the supervised learning. [/INST] So from a practical standpoint, I think that one one thing that is uh you know, potentially a little tough now and this is I think something that we'll see this is a gap that we might see closing over the next couple of years is the ability of reinforcement learning algorithms to effectively utilize large amounts of prior data. So one of the reasons why it's a bit difficult today to use reinforcement learning for all the things that we might want to use it for is that in most of the settings where we want to do rational decision making? Um It's a little bit tough to just deploy some policy that does crazy stuff and learns purely through trial and error. It's much easier to collect a lot of data, a lot of logs of some other policy that you've got and then maybe you you know if you can get a good policy out of that then you deploy it and let it kind of fine tune a little bit. But algorithmic li it's quite difficult to do that. So I think that once we figure out how to get reinforcement learning to bootstrap effectively from large data sets then we'll see a very very rapid growth in applications of these technologies. So this is what's referred to as off policy reinforcement learning or offline RL or batch RL. Um And I think we're seeing a lot of research right now that does bring us closer and closer to that. </s><s>[INST] Can you maybe paint the picture of the different methods she said? Uh Off policy, what's value based reinforcement learning, what's policy based, what's model based, what's off policy on policy, what are the different categories of [/INST] reinforcement? So one way we can think about reinforcement learning is that it's it's in some very fundamental way it's about learning models that can answer kind of what if questions So what would happen if I take this action that I hadn't taken before? Um and you do that of course from experience from data and often times you do it in a loop. So you build a model that answers these, what if questions uh use it to figure out the best action you can take and then go and try taking that and see if the outcome agrees with what you predicted. Um So uh the different kinds of techniques basically refer different ways of doing. And so a model based methods answer a question of uh what state you would get, basically what would happen to the world if you would take a certain action. Value based methods? Uh they answer the question of what value you would get, meaning, what utility you would get. Um but in a sense they're not really all that different because they're both really just answering these. What if questions now, Unfortunately for us with current machine learning methods answering what if questions can be really hard because they are really questions about things that didn't happen if you want to answer. What if questions about things that did happen, you wouldn't need to learn model, you would just like repeat the thing that worked before. Um and that's really a big part of why RL is a little bit tough. So if you have a purely on policy kind of online process then you ask these what if questions you make some mistakes, then you go and try doing those mistaken things and then you observe kind of the counter examples that will teach you not to do those things. Again if you have a bunch of off policy data and you just want to synthesize the best policy you can out of that data then you really have to deal with the challenges of making these these counterfactual. </s><s>[INST] First of all, what's the policy? [/INST] A policy is uh a model or some kind of function that maps from observations of the world to actions. So in reinforcement learning, we often refer to the uh the current configuration of the world as the state. So we say the state kind of encompasses everything you need to fully define where the world is at at the moment. And depending on how we formulate the problem we might say you either get to see the state or you get to see an observation which is some snapshots and peace of the state. </s><s>[INST] So policy is just includes everything in it in order to be able to act in this world. And so what does off policy mean [/INST] if Yeah. So the terms on policy and off policy refer to how you get your data. So if you get your data from somebody else who was doing some other stuff, maybe you get your data from some manually programmed uh system that was you know just running in the world before that's referred to as off policy data. But if you got the data by actually acting in the world based on what your current policy thinks, it's good. We call that on policy data and obviously on policy data is more useful to you because if your current policy makes some bad decisions, you will actually see that those decisions are bad off policy data however might be much easier to obtain because maybe that's all the log data that you have from before. </s><s>[INST] So we talked about offline talked about autonomous vehicles. So you can envision off policy kind of approaches in uh robotic spaces where there's already a ton of robots out there, but they don't get the luxury of being able to uh explore based on reinforcement learning framework. So how do we make again? Open question, But how do we make our policy methods work? [/INST] Yeah. So uh this is something that has been kind of a big open problem for a while and in the last few years uh people have made a little bit of progress on that. Um You know, I can tell you about and it's not by any means solved yet, but I can tell you some of the things that for example we've done to try to address some of the challenges. Um It turns out that one really big challenge with off policy reinforcement learning is that you can't really trust your models to give accurate predictions for any possible action. So if I've never tried to if in my data set, I never saw somebody steering the car off the road onto the sidewalk, my value function or my model is probably not going to predict the right thing if I asked what would happen if I were to steer the car off the road onto the sidewalk. So one of the important things you have to do to get off policy RL to work is you have to be able to figure out whether a given action will result in a trustworthy prediction or not. And you can use kind of distribution uh estimation methods, kind of density estimation methods to try to figure that out. So you could figure out that Well, this action, my model is telling me that it's great, but it looks totally different from any action taken before. So my model is probably not correct. Uh and you can incorporate regularization terms into your learning objectives, that will essentially tell you not to ask those questions that your model is unable to answer, </s><s>[INST] What would lead to breakthroughs in this space. Do you think? Like what's needed? Is this a dataset question? [/INST] Do we need to </s><s>[INST] collect big benchmark data sets that allow us to explore the space? Is that a new kinds of methodologies like what's your sense? Or maybe coming together in a space of robotics and defining the right problem to be working on, [/INST] I think for off policy reinforcement in particular, it's very much an algorithms question right now And uh you know, this is something that I think it's great because alexander's question is, you know, that that just takes some very smart people to get together and think about it really hard. Whereas if it was like a data problem or a hardware problem, that would take some serious engineering. So that's why I'm pretty excited about that problem because I think that we're in a position where we can make some real progress on it just by coming up with the right algorithms in terms of which algorithms that could be, you know, the problems at their core are very related to problems in, you know, things like, like causal inference, right? Because what you're really dealing with the situations where you have a model of statistical model that's trying to make predictions about things that I hadn't seen before. Uh and if it's a if it's a model is generalizing properly, that'll make good predictions. If it's a model that picks up on spurious correlations that will not generalize properly. And then you can you have an arsenal of tools you can use, you could for example, figure out what are the regions where is trustworthy or on the other hand, you could try to make it generalist better somehow or some combination of the two, </s><s>[INST] is there a room for mixing Sort of where most of it like 1995% is off policy. You already have the data set and then you get to send the robot out to do a little exploration like what was that role of mixing them together. [/INST] Yeah, absolutely. I think that this is something that you actually uh described very well at the beginning of the of our discussion, when you talked about the iceberg like this is the iceberg that the 99% of your prior experience, that's your iceberg, you use that for off policy reinforcement learning. And then of course if you've never Uh you know, opened that particular kind of door with that particular lock before, then you have to go out and fiddle with it a little bit and that's that additional 1% to help you figure out a new task. And I think that's actually like a pretty good recipe going forward. </s><s>[INST] Is this to you the most exciting space of reinforcement learning now or is there what's uh maybe taking a step back? Not just now, but what's to use the most beautiful idea, apologize for the romanticized question, but the beautiful idea or concept in reinforcement learning [/INST] in general. I actually think that one of the things that is a very beautiful idea in reinforcement learning is just the idea that you can uh obtain a near optimal controller in your optimal policy without actually having a complete model of the world. This is um you know, so it's something that feels perhaps kind of obvious uh if you if you just hear the term reinforcement learning or you think about trial and error learning but from a controls perspective, it's a very weird thing because classically uh you know, we we think about engineered systems and controlling engineered systems as uh as the problem of writing down some equations and then figuring out given these equations, you know, basically solve for x, figure out the thing that maximizes its performance. And the theory of reinforcement learning actually gives us a mathematically principled framework to think to reason about, you know, optimizing some quantity when you don't actually know the equations that govern that system. And that to me that's actually seems kind of kind of, you know, very elegant, not not something that uh sort of becomes immediately obvious, at least in the mathematical sense. </s><s>[INST] Does it make sense to you that it works at all? [/INST] Well, I think it makes sense when you take some time to think about it, but it is a little surprising. </s><s>[INST] Well then uh then taking a step into the more deeper representations, which is also very surprising of sort of the richness of the state space, the space of environments that this kind of approach can operate. And can you maybe say what is deep reinforcement learning? [/INST] Well, deep reinforcement learning simply refers to taking reinforcement learning algorithms and combining them with high capacity neural net representations, um which is you know, kind of it might at first seem like a pretty arbitrary thing, just take these two components and stick them together. But the reason that it's um it's something that has become so important in recent years is that reinforcement learning, it kind of faces an exacerbated version of a problem that has faced many other machine learning techniques. So if you if we go back to like, you know, the early two thousand's or the late nineties, we'll see a lot of research on machine learning methods that have some very appealing mathematical properties. Like they reduce the convex optimization problems for instance, but they require very special inputs. They require a representation of the input that is clean in some way, like for example, clean in the sense that the classes in your multi class classification problem, separate linearly. So they have some kind some kind of good representation. And we call this feature representation. Uh and for a long time people were very worried about features in the world of supervised learning because somebody had to actually build those features. So you couldn't just take an image and plug it into your logistic regression or your Spm or something. Someone had to take that image and process it using some handwritten code. And the neural nets came along and they could actually learn the features and suddenly we could apply learning directly to the raw inputs, which was great for images. But it was even more great for all the other fields where people hadn't come up with good features yet. And one of those fields actually reinforcement learning because in reinforcement learning, the notion of features, if you don't use neural nets and you have to design your own features is very, very opaque. Like it's very hard to imagine. Let's say I'm playing chess or go what is a feature with which I can represent the value function for go or, or even the optimal policy for go linearly. Like I don't, I don't even know how to start thinking about it and, and people try all sorts of things that would write down, you know, an expert chess player looks for whether the night is in the middle of the board or not. So that's a feature is night in middle of board and they would write these like long lists of kind of arbitrary made up stuff and that was really kind of getting a snowboard, </s><s>[INST] but, and that's a little chess is a little more accessible than the robotics [/INST] problem. Absolutely. </s><s>[INST] Right. That there's at least experts in the different features for chess, [/INST] but still </s><s>[INST] like the neural network there. Uh, to me that's uh, I mean you put it eloquently and almost made it seem like a natural step to add neural networks, but the fact that your networks are able to discover features in the control problem. It's very interesting. It's hopeful, I'm not sure what to think about it, but it feels hopeful that the control problem has features to be learned? Like I guess my question is is it surprising to you how far the deep side of deep reinforcement learning was able to like what the space of problems has been able to tackle from especially in games with the Alpha star and and uh Alpha zero and just the representation power there and in the robotic space. And what what what is your sense of the limits of this representation power and the control context? [/INST] I think that in regard to the limits a bit here, I think that one thing that makes it a little hard to fully answer this question is because um in settings where we would like to push these things to the limit we encounter other bottlenecks. So like the reason that I can't get my robot to learn how to like I don't know do the dishes in the kitchen. It's not because it's neural net is not big enough. It's because when you try to actually do trial and error learning reinforcement learning directly in the real world where you have the potential to gather these large uh there you know highly varied and complex data sets. You start running into other problems like one problems you run into very quickly. Uh It'll first sound like a very pragmatic problem that actually turns out to be a pretty deep scientific problem. Take the robot put in your kitchen have to try to learn to do the dishes with trial and error. It will break all your dishes and then we'll have no more dishes to clean. Now you might think this is a very practical issue, but there's something to this which is that if you have a person trying to do this, you know a person will have some degree of common sense, they'll break one dish will be a little more careful with the next one and if they break all of them they're going to go and get more or something like that. So there's all sorts of scaffolding that that comes very naturally to us for our learning process. Like you know if I have to learn something through trial and error, I have the common sense to know that I have to try multiple times if I screw something up, I ask for help or I reset things or something like that and all of that is kind of outside of the classic reinforcement problem formulation. Uh There are other things that are uh that can also be categorized as kind of scaffolding but are very important, like for example, where do you get your reward function if I want to learn how to pour a cup of water? Uh Well how do I know if I've done it correctly now, that probably requires an entire computer vision system to be built just to determine that and that seems a little bit inelegant. So there are all sorts of things like this that start to come up when we think through what we really need to get reinforcement learning to happen at scale in the real world and that many of these things actually suggest a little bit of a shortcoming in the problem formulation and a few deeper questions that we have to resolve. </s><s>[INST] That's really interesting. I talked to like David Silver bought off a zero and it seems like there's no, again, the we haven't hit the limit at all uh in the context when there is no broken dishes. So in the, in the case of go, you can, it's really about just scaling compute. So, again, like the bottleneck is the amount of money you're willing to invest in compute and then maybe the different the scaffolding around how difficult it is to scale compute maybe, but there there's no limit. And it's interesting now we move to the real world and there's the broken dishes, there's all the and the reward function we can mention. That's really nice. So, what, how do we push forward there? Do you think there's, there's this kind of sample efficiency question that people Bring up, you know, not having to break 100,000 dishes. Is this an algorithm question, is this a data selection like question, what do you think? How do we how do we not break too many dishes? [/INST] Yeah. Um well, one way we can think about that is that um maybe we need to be better at at reusing our data building that that iceberg. So perhaps perhaps it's too much to hope that you can have a machine that in isolation in the vacuum without anything else, can just master complex tasks in like in minutes, the way that people do, but perhaps it also doesn't have to, perhaps what it really needs to do is have an existence a lifetime where it does many things and the previous things that it has done, prepare it to do new things more efficiently. And you know, the, the study of these kinds of questions typically falls under categories like multitask learning or metal learning, but they all fundamentally deal with the same general theme, which is use experience for doing other things to learn to do new things efficiently and quickly. </s><s>[INST] So what do you think about if you just look at the one particular case study of Tesla autopilot that has quickly approaching towards a million vehicles on the road where some percentage of the time, 30 40% of the time is driving using the computer vision, Multitask hydra net. Right. Uh and then the other percent, that's what they call it, hydro net. Uh the other percent is human controlled from the human side. How can we use that data? What's your sense? So like what's the signal? Do you have ideas in this autonomous vehicle space when people can lose their lives? You know, it's a it's a safety critical environment. So how do we use that data? [/INST] So, um I think that actually the kind of problems that come up when we want um systems that are reliable and that can kind of understand the limits of their capabilities. They're actually very similar to the kind of problems that come up when we were doing off policy reinforcement learning. So as I mentioned before and off policy reinforcement learning, the big problem is you need to know when you can trust the predictions of your model because if you if you're trying to evaluate some pattern of behavior for which your model doesn't give you an accurate prediction, then you shouldn't use that to modify your policy. And it's actually very similar to the problem that we're faced when we actually then deploy that thing and we want to decide whether we trust it in the moment or not? Um So perhaps we just need to do a better job of figuring out that part. And that's a very deep research question of course, but it's also a question that a lot of people are working on. So I'm pretty optimistic that we can make some progress on that over the next few years. </s><s>[INST] What's the role of simulation in reinforcement learning? Deep reinforcement learning reinforcement learning? Like how essential is it? It's been essential for the breakthroughs so far for some interesting breakthroughs. Do you think it's a crutch that we rely on? I mean, again it connects to our off policy discussion, but do you think we can never get rid of simulation? Or do you think simulation will actually take over will create more and more realistic simulations that will allow us to uh to solve actual real world problems. Like transfer the models who learning simulation from real world problems. [/INST] I think that simulation is a very pragmatic tool that we can use to get a lot of useful stuff to work right now. But I think that in the long run we will need to build machines that can learn from real data because that's the only way that will get them to improve perpetually. Because if we can't have our machines learn from real data, if they have to rely on simulated data, eventually the simulator becomes the bottleneck. In fact, this is a general thing. If your machine has any bottleneck that is built by humans and that doesn't improve from data, it will eventually be the thing that holds it back. Uh And if you're entirely reliant on your simulator, that will be the bottleneck. If you're entirely reliant on a manually designed controller that's going to be the bottleneck. So, simulation is very useful. It's very pragmatic, but it's not a substitute for being able to utilize real experience. Uh And this is by the way, this is something that I think is quite relevant now, especially in the context of some of the things we've discussed? Because some of these kind of scaffolding issues that I mentioned, Things like the broken dishes and the unknown reward function, like these are not problems that you would ever stumble on uh when working in a purely simulated kind of environment. But they become very apparent when we try to actually run these things in the real world </s><s>[INST] to throw a brief french in our discussion. Let me ask, do you think we're living in a simulation? [/INST] Oh, I have no idea. </s><s>[INST] Do you think that's a useful thing to even think about about the the fundamental physics nature of reality? Or another perspective? Reason, I think the simulation hypothesis interesting is is to think about how difficult is it to create sort of a virtual reality game type situation that will be sufficiently convincing to us humans or sufficiently enjoyable that we wouldn't want to leave. That's actually a practical engineering challenge. And I I personally really enjoy virtual reality but it's quite far away. But I kind of think about what would it take for me to want to spend more time in virtual reality versus the real world. And that's a that's a sort of a nice clean question because at that point we reached, if I want to live in a virtual reality, that means we're just a few years away where a majority of the population lives in a virtual reality and that's how we create the simulation, right? You don't need to actually simulate the quantum gravity and just every aspect of the of the universe. And that's a that's an interesting question for reinforcement learning to is if we want to make sufficiently realistic simulations that made it blends the difference between sort of the real world and the simulation thereby just some of the things we've been talking about kind of the problems go away. If we can create actually interesting rich simulations. [/INST] It's an interesting question and it actually I think your question casts your previous question to in a very interesting light because in some ways asking whether we can um well the more the more kind of practical version of this like you know can we build simulators that are good enough to train essentially ai systems that will work in the world. And it's kind of interesting to think about this about what this implies. If true, it kind of implies that it's easier to create the universe than it is to create a brain. And that seems like put this way it seems kind of weird. </s><s>[INST] The aspect of the simulation most interesting to me is the simulation of other humans that seems to be mhm. A complexity that makes the robotics problem harder. Now, I don't know if every robotics person agrees with that notion. Just as a quick aside what are your thoughts about when the human enters the picture of the robotics problem? How does that change the reinforcement learning problem? The learning problem in general. [/INST] Yeah, I think that's a it's a kind of a complex question and I guess my hope for a while had been that if we build these robotic learning systems that are multi task that utilize lots of prior data and that learn from their own experience, the bit where they have to interact with people will be perhaps handled in much the same way as all the other bits. So if they have prior experience of interacting with people and they can learn from their own experience of interacting with people for this new task, maybe that'll be enough. Um Now of course there if it's not enough, there are many other things we can do and there's quite a bit of research in that area but I think it's worth a shot to see whether the uh the multi agent interaction, the ability to understand that other beings in the world have their own goals, intentions and thoughts and so on. Whether that kind of understanding can emerge automatically from simply learning to do things with and maximize utility. </s><s>[INST] That information arises from the data. You said something about gravity sort of uh that you don't need to explicitly inject anything into the system. They can be learned from the data and gravity is an example of something that could be learned from data sort of like the physics of the world. Like what what are the limits of what we can learn from data? Do you really do you think we can So a very simple, clean way to ask that is, do you really think we can learn gravity from just data? The idea of the laws of gravity? [/INST] So this is something that I think is is a common um kind of pitfall when thinking about prior knowledge and learning is to assume that just because we know something, then it's better to tell the machine about that rather than have it figured out and so on. In many cases, things that are important that affect many of the events that the machine will experience are actually pretty easy to learn. Like, you know, if things, if every time you drop something it falls down, like Yeah, you might not get the, you know, you might get kind of Newton's version, not Einstein's version, but will be pretty good and it will probably be sufficient for you to act rationally in the world because you see the phenomenon all the time. So, things that are readily apparent from the data, we might not need to specify those by hand, it might actually be easier to let the machine figure them out. </s><s>[INST] It just feels like that there might be a space of many local, a local minimum in terms of theories of this world that we would discover and get stuck on. That Newtonian mechanics is not necessarily easy to come by. [/INST] Yeah, and well, in fact, in in some fields of science, for example, human civilizations, that cell full of these local optima. So for example, if you think about how people uh tried to figure out biology and medicine, you know, for the longest time, the kind of rules like the kind of principles that serve us very well in our day to day lives actually serve us very poorly in understanding medicine and biology. We had kind of very uh superstitious and weird ideas about how the body worked until the advent of the modern scientific method. So that does seem to be, you know, a failing of this approach, but it's also a failing of human intelligence arguably </s><s>[INST] maybe a small side, but some, you know, the idea of self play is fascinating and reinforcement learning sort of these competitive, creating a competitive context in which agents can play against each other in a at the same skill level and thereby increasing each other's skill of. It seems to be this kind of self improving mechanism is exceptionally powerful in the context where it could be applied first of all said beautiful to you, that this mechanism work as well as it does and also can be generalized to other context, like in the robotic space or anything that's applicable to the real world. [/INST] I think that um it's a very interesting idea and I but I suspect that the bottleneck to actually generalizing it to the robotic setting is actually gonna be the same as as the bottleneck for everything else that we need to be able to build, machines that can get better and better through natural interaction with the world. And once we can do that then they can go out and play with, they can play with each other, they can play with people, they can play with the natural environment. Uh But before we get there we've got all these other problems we've got we have to get out of the way. </s><s>[INST] There is no shortcut around that you have to interact with the natural environment that [/INST] well because in a in a self place setting you still need a mediating mechanism. So the reason that um you know self play works for a board game is because the rules of that board game mediate the interaction between the agents. So the kind of intelligent behavior that will emerge depends very heavily on the nature of that mediating mechanism. </s><s>[INST] So on the side of reward functions that's coming up with good reward function seems to be the thing that we associate with general and like human beings seem to value the idea of developing our own reward functions of uh you know arriving at meaning and so on. And yet for reinforcement learning we often kind of specified as the given. What's your sense of how we develop reward again you know good reward functions. [/INST] Yeah I think that's a very complicated and very deep question and you're completely right that classically in reinforcement learning this question. Yes kind of been treated as a non issue, that you sort of treat the reward as this external thing that comes from some other bit of your biology and you kind of don't worry about it. And I do think that that's actually you know, a little bit of a mistake that we should worry about it. And we can approach in a few different ways we can approach it for instance by thinking of rewards as a communication medium, we can say well how does a person communicate to a robot, what its objective is? You can approach it also as a sort of more of an intrinsic motivation meeting you could say, can we write down kind of a general objective that leads to uh good capability? Like for example, can you write down some objective, such that even in the absence of any other task, if you maximize that objective, you sort of learn useful things. Um This is something that has sometimes been called unsupervised reinforcement learning, which I think is a really fascinating area of research, especially today, we've done a bit of work on that recently. One of the things we've studied is whether we can have some notion of uh of unsupervised reinforcement learning by means of um you know, information theoretic quantities like for instance minimizing a Beijing measure of surprise. This is an idea that was uh you know pioneered actually in the computational neuroscience community by folks like carl kristen uh and we've done some work recently that shows that you can actually learn pretty interesting skills by essentially behaving in a way that allows you to make accurate predictions about the world. It seems a little circular, like do the things that will lead to you getting the right answer uh for for prediction, But you can, you know, by doing this, you can sort of discover stable niches in the world, You can discover that if you're playing Tetris then correctly, uh you know, clearing the roads, will let you play Tetris for longer and keep the board nice and clean, which sort of satisfies some desire for order in the world. Uh and as a result, to get some degree of leverage over your domain. Um so we're exploring that pretty </s><s>[INST] actively. Is there a role for a human notion of curiosity in itself, being the reward, sort of discovering new things about the the world? [/INST] So one of the things that I'm pretty interested in is actually whether discovering new things can actually be an emergent property of some other objective that quantifies capability. So, new things for the sake of new things, maybe it's not, maybe might not by itself be the right answer, but perhaps we can figure out an objective for which discovering new things is actually the natural consequence? Uh that's something we're working on right now, but I don't have a clear answer for you there yet, that's still work in progress. </s><s>[INST] You mean, just as a psa curious observation to see sort of creative uh patterns of curiosity on the way to optimize for a particular, [/INST] on the way to optimize for a particular measure of capability. </s><s>[INST] Is uh is there ways to understand or anticipate unexpected, unintended consequences of particular reward functions, sort of uh anticipated kind of strategies that might be developed and try to avoid highly detrimental strategies? [/INST] Uh So classically this is something that has been pretty hard in reinforcement learning because it's difficult for a designer to have good intuition about, you know, what a learning album will come up with when they give it some objective. Um There are ways to mitigate that. One way to mitigate it is to actually define an objective that says like don't do weird stuff, you can actually quantify it and say just like don't enter situations that have low probability under the distribution of states you've seen before. It turns out that's actually one very good way to do off policy reinforcement learning actually. Um So we can do some things like that. </s><s>[INST] If we uh slowly venture in speaking about reward functions into greater and greater levels of intelligence, there's uh means to Russell thinks about this, the alignment of Ai systems with us humans. So how do we ensure that g systems align with those humans? It's a it's kind of a reward function. Question of specifying the behavior of ai systems such that their success aligns with the broader intended success interest of human beings, you have thoughts on this? They have kind of concerns of where reinforcement learning fits into this? Are you really focused on the current moment of us being quite far away and trying to solve the robotics problem? [/INST] I don't have a great answer to this, but um you know, and I do think that this is a problem that's that's important to figure out for my part. I'm actually a bit more concerned about the other side of the of this equation that, you know, maybe rather than unintended consequences for uh objectives that are specified to. Well, I'm actually more worried right now about unintended consequences for objectives that are not optimized well enough, uh which might become a very pressing problem when we for instance, try to use these techniques for safety critical systems like cars and aircraft and so on. I think at some point we'll face the issue of objectives being optimized too well, but right now, I think we're more likely to face the issue of them not being optimized well enough, </s><s>[INST] but you don't think unintended consequences can arise even when you're far from optimal quality, sort of like on the path to it. [/INST] Oh no, I think unintended consequences can absolutely arise. Uh It's just, I think right now the bottleneck for improving reliability, safety and things like that is more with systems that need to work better that need to, optimize the objective better, </s><s>[INST] Do you have thoughts concerns about existential threats Of human level intelligence sort of if we put on our head of looking in 10, 20, 100 500 years from now, you have concerns about existential threats of Ai systems. [/INST] I think there are absolutely existential threats for Ai systems, just like there are for any powerful technology. Um but I think that the these kinds of problems can take many forms and uh and some of those forms will come down to uh you know, people with nefarious intent, some of them will come down to A I. Systems that have some fatal flaws and some of them will will of course come down to A. I. Systems that are too capable in some way. But among this set of potential concerns, I would actually be much more concerned about the first two right now and principally the one with nefarious humans. Because you know, just through all of human history, actually the nefarious humans that have been the problem, not the nefarious machines. Um then I am about the others and I think that right now the best that I can do uh to make sure things go well, is to build the best technology I can and also hopefully promote responsible use of that technology, </s><s>[INST] do you think are all systems has something to teach us humans? You said nefarious humans getting us in trouble. I mean machine learning systems of in some ways have revealed to us the ethical flaws in our data in that same kind of way can reinforcement learning teach us about ourselves. Has it taught something? What have you learned about yourself from trying to build robots and reinforce the learning systems? [/INST] I'm not sure what I've learned about myself, but maybe um part of the answer to your question might become a little bit more apparent once we see more widespread deployment of reinforcement learning for decision making, support in uh you know, in domains like, you know health care, education, social media, etcetera. And I think we will see some interesting stuff emerged there. We will see for instance, what kind of behavior is these systems come up with in situations where there is interaction with humans and where they have the possibility of influencing human behavior. I think we're not quite there yet. But you know, maybe in the next few years we'll see some interesting stuff coming out in that area. </s><s>[INST] I hope outside the research space because the the exciting space where this could be observed is uh sort of a large companies that deal with large data and I hope there's some transparency is and one of the things that's unclear when I look at social networks and just online is why an algorithm did something or whether, you know, even an algorithm was involved. And that would be interesting as a former research perspective just to uh to observe the results of algorithms to open up that data to or to at least be sufficiently transparent about the behavior of these as systems in the real world. What's your sense? I don't know if you looked at the block post bitter lesson by every sudden where it looks at sort of the big lesson of researching ai and reinforced learning. Is that simple methods? General methods that leverage computation seemed to work well. So basically don't try to do any kind of fancy algorithms. Just wait for competition to get fast. Do you share this kind of intuition? [/INST] I think the high level idea makes a lot of sense. I'm not sure that my takeaway would be that we don't need to work on algorithms. I think that my takeaway would be that we should work on general algorithms. Uh And actually I think that this idea of um needing to better automate the acquisition of experience in the real world actually follows pretty naturally from Rich Sutton's conclusion. So if the claim is that automated general methods plus data leads to good results, then it makes sense that we should build general methods and we should build the kind of methods that we can deploy and get them to go out there and like collect their experience autonomously. I think that, you know, one place where I think that the current state of things falls a little bit short of that is actually the the going out there and collecting the data autonomously. Which is easy to do in a simulated board game but very hard to do in the real world. </s><s>[INST] Yeah it keeps coming back to this one problem. Right? It's uh so your mind is focused they're now in this real world. It just seems scary. The step of collecting the data and it seems unclear to me how we can do it effectively. Well you [/INST] know seven billion people in the world. Each of them have to do that at some point in their </s><s>[INST] lives and we should leverage that experience that they've all done the we should be able to try to collect that kind of data. Okay big questions. Maybe stepping back to your life. Would book or books, technical or fiction or philosophical had a big impact on and the way you saw the world and the way you thought about in the world your life in general. Mm And maybe what books if it's different, would you recommend people consider reading on their own intellectual journey? It could be within a reinforcement learning but it could be very much bigger. [/INST] I don't know if this was like a a scientifically like particularly meaningful answer but like the honest answer is that I I actually found a lot of the work by Isaac Asimov to be very inspiring when I was younger. I don't know if that has anything to do with with ai necessarily </s><s>[INST] you don't think it had a ripple effect in your life. [/INST] Maybe it did. Um but yeah, like I think that a vision of a future where um, well, first of all artificial mice, artificial intelligence system, artificial robotic systems have uh, you know, kind of a big place, a big role in society and where uh, we try to imagine the sort of the limiting case of technological advancement and how that might play out in our future history. But yeah, I think that the, that was in some way influential. I don't really know how, but </s><s>[INST] uh, [/INST] I would recommend it. I mean, if nothing else you'd be well entertained, </s><s>[INST] When did you first yourself like fall in love with the idea of artificial intelligence get captivated by this field. [/INST] So my honest answer here is actually that I only really started to think think about it as a, that's something that I might want to do actually in graduate school pretty late. And a big part of that was that, until you know somewhere around 2009, it just wasn't really high on my priority list because I didn't think that it was something where we're going to see very substantial advances in my lifetime. Um and uh, you know, maybe in terms of my career at the time when I really decided I wanted to work on this was when I actually took a seminar course that was taught by professor Andrew ng and uh, you know at that point, of course had some had like a decent understanding of the technical things involved. But one of the things that really resonated with me was when he said in the opening lecture, something to the effect of like, well he used to have graduate students come to him and talk about how they want to work on ai and he would kind of chuckle and give them some math problem to deal with. But now he's actually thinking that this is an area where we might see like substantial advances in our lifetime. And that kind of got me thinking because, you know, in some abstract sense, Yeah, like you can kind of imagine that, but in a very real sense when someone who had been working on the kind of stuff their whole career suddenly says that yeah, like that, that had some effect on me. </s><s>[INST] Yeah, this might be a special moment in the history of the field that this is uh well we might see some some interesting breakthroughs. So in the space of advice, somebody who's interested in getting started in machine learning or reinforcement learning, what advice would you give to maybe an undergraduate student or maybe even younger, how, what are the first steps to take and further on, what are the States steps to take on that journey. [/INST] So something that I think is important to do is to um it's to not be afraid to like spend time imagining the kind of outcome that you might like to see. So, you know, one outcome might be a successful career, um, a large paycheck or something or state of the art result in some benchmark. But hopefully that's not the thing that's like the main driving force for somebody. But I think that if someone who is a student considering a career in ai like takes a little while, sits down and things like what do I really want to see what I want to see a machine do what I want, what do I want to see a robot do what I want to do and what I want to see a natural language system. Just like imagine, you know, imagine it almost like a commercial for a future product or something or like, like something that you'd like to see in the world and then actually sit down and think about the steps that are necessary to get there. And hopefully that thing is not a better number on image net classification. It's like, it's probably like an actual thing that we can't do today. That would be really awesome, whether it's a robot butler or a uh, you know, a really awesome healthcare decision making supports us and whatever it is that you find inspiring. And I think that thinking about that and then backtracking from there and imagining the steps needed to get there will actually too much better research. It will lead to rethinking the assumptions it will lead to working on the bottle necks and other people aren't working on </s><s>[INST] and then naturally returned to you, we've talked about reward functions. You've just given advice on looking forward how you would like to see what kind of change you would like to make in the world. What do you think? Ridiculous big question. What do you think is the meaning of life? What is the meaning of your life? What gives you fulfillment purpose, happiness and meaning? [/INST] That's a very big question. Um </s><s>[INST] what's the reward function under which you are operating? [/INST] Yeah. One thing that does give, you know, if not meaning at least satisfaction is uh some degree of confidence that I'm working on a problem that really matters. I feel like it's less important to me to like actually solve a problem. But it's quite nice to uh take things to spend my time on that I believe really matter. And you know, I I tried pretty hard to look for that. </s><s>[INST] I don't know if it's easy to answer this, but if you're successful, what does that look like? What's the big dream now? Of course success is built on top of success and you keep going forever. But what okay, is the dream? [/INST] Yeah, so one very concrete thing or maybe as concrete as it's gonna get here is is to see machines that actually get better and better the, you know, the longer they exist in the world and that kind of seems like on the surface, one might even think that that's something that we have today, but I think we really don't. I think that there is uh an ending complexity in the universe and to date all of the machines that we've been able to build, don't sort of improve up to the limit of that complexity. They hit a wall somewhere, Maybe they hit a wall because they're in a simulator that has, that is only a very limited, very pale imitation of the real world or they hit a wall because they rely on a labeled dataset, but they never hit the wall of like running out of stuff to see like the uh so you know, I'd like to build a machine that can go as far as possible and that </s><s>[INST] runs up against the ceiling of the complexity of the universe. Yes, well I don't think there's a better way to end it. Sergei thank you so much. It's a huge honor. I can't wait to see the amazing work that you have to publish and in education space in terms of reinforcement learning. Thank you for inspiring the world. Thank you for the great research you do. Thank you. Thanks for listening to this conversation with Sergey Levine and thank you to our sponsors. Cash up and Express VPN, please consider supporting this podcast by downloading cash app and using code lex podcast and signing up at express VPN dot com slash lex pod, click all the links by all the stuff that's the best way to support this podcast and the journey I'm on. If you enjoy this thing, subscribe on Youtube, review it with five stars on apple podcast, support on Patreon or connect with me on twitter at lex Friedman, spelled somehow. If you can figure out how without using the letter E, just F R I D M A N and now let me leave you with some words from Salvador dali. Intelligence without ambition is a bird without wings. Thank you for listening and hope to see you next time. [/INST] </s>

<s>[INST] Do you have to start at the very beginning or can you do a shortcut? [/INST] We're </s><s>[INST] growing a hand. You already got the whole organism. Yeah. [/INST] So here's what we've done. Right? So, so we've we've more or less solved that in frogs, frogs and salamanders do not regenerate their legs as adults. And so so, uh we've shown that with a very uh kind of simple intervention. So what we do is there's two things you need to uh you need to have a signal that tells the cells what to do and then you need some way of delivering it. And so this is work together with um with David Kaplan and I should do a disclosure here. We have a company called morpheus Ooh, tickles and Spinoff where we're trying to to address regenerated, you know, limb regeneration. So we've solved it in the frog and we're now in trials in mice. So now we're going to we're in mammals. Now I can't say anything about how it's going. But the frog thing is solved. So what you do is after </s><s>[INST] a little frog blue sky walker? Yeah, [/INST] basically, basically, yeah. Yeah. So what you do is we did, we did with legs instead of farms and what you do is after amputation. Normally they don't regenerate, you put on a wearable bio reactor. So it's this thing that that goes on and Dave Kaplan, Those lab makes these things and inside it's a, it's a very controlled environment. It is a silk gel that carries some drugs, for example, ion channel drugs. And what you're doing is you're saying to these cells, you should re grow what normally goes here. So, uh, that whole thing is on for 24 hours and then you take it off, you don't touch the link again. This is really important because what we're not looking for is a set of micro management, you know, printing or controlling the cells we want to trigger, we want to we want to interact with it early on and then not touch it again because because we don't know how to make a frog leg, but the frog knows how to make a frog leg. So 24 hours, 18 months of like growth after that without us touching it again. And after 18 months you get a pretty good leg. That kind of shows this proof of concept that early on when the cells right after injury, when they're first making a decision about what they're gonna do, you can you can impact them and once they've decided to make a leg, they don't need you after that they can, you know, do their own thing. So that's an approach that we're now taking. </s><s>[INST] What about cancer suppression? That's something you mentioned earlier. How can all of these ideas help with cancer suppression? So [/INST] let's let's go back to the beginning and ask what what what what cancer is. So I think um you know, asking why there's cancer is the wrong question. I think the right question is why is there ever anything but cancer? So so in the normal state you have a bunch of cells that are all co cooperating towards a large scale goal. If that process of cooperation breaks down and you've got a cell that is isolated from that electrical network that lets you remember what the big goal is. You revert back to your uni cellular lifestyle as far as I think about that border between self and world, right? Normally when all these cells are connected by gap junctions into an electrical network, they are all one self right? Meaning that their goals, they have these large tissue level goals and so on. As soon as the cell is disconnected from that, the self is tiny. Right? And so at that point. And so so people, a lot of people model cancer cells cells as being more selfish and all that they're not more selfish. They're equally selfish. It's just that their self is smaller. Normally the self is huge. Now they got tiny little cells. Now, what are the goals of tiny little selves well proliferate and migrate to wherever life is good and that's metastasis that's proliferation and metastasis. So so one thing we found and people have noticed years ago that when cells convert to cancer, the first thing they see is they close the gap junctions and it's a lot like I think it's a lot like that experiment with the slime mold where until you close that gap junction you can't even entertain the idea of leaving the collective because there is no you at that point right? Your mind melded with this with this whole other network. But as soon as the gap junction is closed now the boundary between you now now the rest of the body is just outside environment to you. You're just you're just a uni cellular organism and the rest of the body's environment. So so we so we studied this process and we worked out a way to artificially control the by electric state of these cells to physically force them to remain in that network. And so then then what that what that means is that um nasty um mutations like a race and things like that. Isn't this really tough psychogenic mutations that caused tumors if you if you do them and then but but then then artificially um uh control of the Buy electrics. You you you you greatly reduced neurogenesis or or normalized cells that had already begun to convert you basically they go back to being normal cells. And so this is another much like with the plan area. This is another way in which the bioelectric state kind of dominates what the what the genetic state is. So if you sequence the you know if you sequence the nucleic acid you'll see the keras mutation. You'll say well that's gonna be a tumor but there isn't a tumor because because by electrically you've kept the cells connected and they're just working on making nice skin and kidneys and whatever else. So so we've started moving that to um to to human glioblastoma cells. And we're hoping for you know a patient in the future interaction with patients. </s>

<s>[INST] to interesting things. You mentioned one is you really have to know what you're talking about, how you've uh, you worked on a lot of very challenging technical things. So I kind of assumed you were born uh, technically savvy, but assuming that's not the case, how do you, how do you develop technical expertise? Like even at google, you work on, I don't know how many projects but really challenging very varied [/INST] compilers, tps hardware, cloud stuff, a bunch of different things. Um, the thing that I've become comfortable is I've more comfortable with as I've gained experience is, uh, being okay with not knowing. And so a major part of leadership is actually, it's not about having the right answer, it's about getting the right answer. And so if you're working in a team of amazing people, right? And many of these places made these companies all have amazing people. It's a question of how do you get people together? How do you get, how do you build trust? How do you get people to open up? How do you people get people to, you know, be vulnerable sometimes with an idea that maybe isn't good enough, but it's the start of something beautiful. How do you, um, how do you provide an environment where you're not just like top down the shell, do the thing that I tell you to do right? But you're encouraging people to be part of the solution and uh, providing a safe space where if you're not doing the right thing, they're willing to tell you about it. So </s><s>[INST] you're asking dumb questions. Yeah. The questions [/INST] are my specialty. Well, so I've been in the harbor recently and I don't know much at all about how chips are designed. I know a lot about using them. I know some of the principles and the arts technical level of this, but it turns out it turns out that if you ask a lot of dumb questions you get smarter really, really quick and when you're surrounded by people that want to teach and learn themselves can be a beautiful thing. </s><s>[INST] So, let's talk about programming languages if it's okay at the highest absurd philosophical level because I don't get [/INST] romantic on me last, </s><s>[INST] I will forever give them, I think, and torture. I apologize. Uh, why do programming languages even matter? [/INST] Okay, well, thank you very much. You're saying why should why why should you care about anyone programming language? Or why do you why do we care about programming computers </s><s>[INST] or no? Why? Why do we, why do we care about programming language? Design? Creating effective programming languages? Uh, choosing a, you know, one program languages versus another programming language. Why we keep struggling and improving through the evolution of these programming languages? [/INST] Okay, so, I mean, I think you have to come back to, what are we trying to do here? Right. So, we have these, these beasts called computers that are very good at specific kinds of things and we think it's useful to have them do it for us. Right? Uh, now you have this question of how best to express that because you have a human brain still has an idea on its head and you want to achieve something. So, well, there's lots of ways of doing this. You can go directly to the machine and speak assembly language and then you can express directly with the computer understands that's fine. Um, you can then have higher and higher and higher levels of abstraction. Up until machine learning and you're designing a neural net to do the work for you. Um, the question is, where, where along this way do you want to stop? And what benefits do you get out of doing so? And so programming languages in general, you have sea of FORTRAN and java and Ada, pascal swift, you have lots of different things. Um, they'll have different tradeoffs and they're tackling different parts of the problems. Now, one of the things that most programming languages do is they're trying to make it so that you have a pretty basic things like portability across different hardware. So you've got, I'm gonna run on an Intel PC. I'm gonna risk five PC. I'm gonna run on arm phone or something like that. Fine. Um, I want to write one program and have it portable. And this is something that assembly doesn't do now when you start looking at the space of programming languages. This is where I think it's fun because programming languages all have tradeoffs and most people will walk up to them and they look at the surface level of syntax and say, oh, I like curly braces, right? Like tabs or I like, you know, semicolons or not or whatever. Right subjective, fairly subjective, very shallow things. But programming languages when done right, can actually be very powerful and the benefit they bring is expression okay, and if you look at programming languages, there's really kind of two different levels to them. One is the down in the dirt nuts and bolts of how do you get the computer to be efficient stuff like that, How they work, type systems, compilers, stuff, things like that. The other is the UI and the UI for programming language is really a design problem and a lot of people don't think about it that way. </s><s>[INST] And the Ui, you mean all that stuff with the braces and yeah, all that [/INST] stuff to you. I and what it is and you, I mean is user interface. Um, and so what what's really going on is it's the interface between the guts and the human and humans are hard. Humans have feelings, they have things they like, they have things they don't like. And a lot of people treat programming languages as though humans are just kind of abstract creatures that cannot be predicted. But it turns out that actually there are, there is better and worse. Like people can tell when a program language is good or when it was an accident, right? And one of the things was swift in particular is that a tremendous amount of time by a tremendous number of people have been put into really polishing and make it feel good. But it also has really good. Nothing bolts underneath it. </s>

<s>[INST] know, [/INST] for people who know they know that it's like Carhartt, even though the Carhartt tag isn't there anymore, you just know by the shape of it, you feel it so it it abstracts, it takes into like this cubist place like Picasso, Matisse or african cubism or dire straits money for nothing and your chicks for free. Look at that. Yeah, </s><s>[INST] I wouldn't think you could reference dire straits [/INST] that video. </s><s>[INST] I know it's a great video [/INST] that your </s><s>[INST] encyclopedic knowledge of music is [/INST] great. I think that, you know, porn is not just sexual. That term actually relates to torture. Like what I'm saying, like watching people beat each other up on, on world Star or something? You know, it's like, that's called torture porn. </s><s>[INST] There's also food porn. How do you explain that [/INST] food porn? </s><s>[INST] Food porn is like looking at really delicious cakes and cookies and steak and so on and you're like holy sh it [/INST] Exactly! Architecture like this. People who love architecture when they see architecture. I know I got friends like oh that's porn. Yeah. Architecture. </s><s>[INST] Yeah. Yeah, it's like uh that's like love like dopamine fueled love for a thing. Yeah, we need that's beautiful. I think that's beautiful. [/INST] Well I think </s><s>[INST] as long as nobody's getting hurt, [/INST] the intoxication that people have, you know, the worst part is like when one person is in love and the other person's didn't like </s><s>[INST] that's us right now [/INST] who's unlike, </s><s>[INST] because you said you love Bill Maher and you like me [/INST] not and and </s><s>[INST] I love you. So that was [/INST] no, I actually love you both. All right. And the funny thing is like um as a as a visionary visionaries are more extreme than business people because a visionary would do bad business for their vision. A business person will never do bad business. It's </s><s>[INST] well put, Yeah, it's really well put. So [/INST] when you say, I don't care about money, it's like, yeah, I feel you like I feel that I care in a way that I need that, but I care about the vision that I have to a steve jobs level </s><s>[INST] to the [/INST] point of putting that above My regular human existence, meaning having to go and take all the arrows that I become a soldier of the vision. Because someone could say a wife could say to like say like in the 60s, uh their husband's about to go to war. The wife may not want their husband to go to war because of their Children, but he's going to war because of their Children </s><s>[INST] definitely. Do you regret any aspect of just giving everything to your vision. I mean you've spoken about that saying like I got fucked over in the business side, [/INST] I promised that I had to stop envying </s><s>[INST] cool [/INST] people at their current cool time. If anything, if someone's cool and I say their name a lot, I'm gonna flash the light on them so much that it ages their coolness. It's actually a war technique showered in compliments, </s><s>[INST] you [/INST] Know because that that you to be able to stay at 45 years old to still be the bad kid in class at 45, I'm still in the principal's office because if you look at like say fashion right, if you think of something fashionable it's probably gonna look like the kid in the principal's </s><s>[INST] office, that's [/INST] gonna be the fashion and that's part of the reason why I'm the most influential person and fashion and fashion because I refuse to leave the principal's office if the substitute teacher comes in, I'm gonna say you got a big </s><s>[INST] but [/INST] I'm gonna get </s><s>[INST] back right back to the principal's office [/INST] and people gonna be like yo and that's what you keep saying like who can we talk to because no one can talk to me, no I'm out of control bro, it's just God then me and then fuck everybody, you know I mean that's how that's really how I live my life and I just want to say this one thing about the Children thing I wanted to you know, put that point together, put that idea together. I don't like the word point. I'm really big on like words, words being powerful says the person with the sleepy tweet um so </s><s>[INST] full of contradictions. [/INST] Yeah, I'm human being right. It's like some days I'm like I said, I could go 30 days to be celibate, you know, like really? Like I'm a monk, I'm being like faithful to God and then I go like working out no carbs, like my work weight will fluctuate and I have times, I'm like, I don't feel like being told what to do. So I don't want to work out with a trainer and I just want to drink this Hennessy. But I want to say this thing that I will text kim sometimes I'll say, you know, you may not understand What I'm doing right now, but you will. And that's what I would say to everyone, you may not understand what I'm doing and this gives a little bit, maybe we're like 20% there. But there's still gonna be people, we leave this interview and they were like, I still don't like the way I apologize, I still don't want to say what my involvement would have been to make him lead him to that point and we just want things to be how they are, there's people in the world who want the world to stay as it is and there's people who want to change the world and that's happened throughout time. They say that with Da Vinci halfway into a conversation, he could not suffer fools so much that he would walk away halfway into the conversation and I said, well, that's why it's the Da Vinci code and at the Da Vinci world and every time I go and piss off a group of people or do these things on some end, it's like I could be turning my thing from a potential world vision to a code or I could be opening up conversations with new friends that I may trust someday. </s>

<s>[INST] still arm lock. And so I meant more, not for the submission, but for the control of the game of it that everybody [/INST] needs opportunity to group together </s><s>[INST] to get, there [/INST] is a week, but </s><s>[INST] you also have groups in nog can't use those [/INST] groups. No, </s><s>[INST] there's a thought experiment. So like I'm trying to understand how essential [/INST] panic </s><s>[INST] panic of course everyone panics. A bear panics when they're in the water with the shark. But that doesn't mean the bear can still win when it stops panicking and relaxes. [/INST] Its not possible. </s><s>[INST] That's another discussion. Can the bear bee the shark in the water? Um actually I need to put maybe a polar bear because they're pretty good at swimming. Okay. [/INST] I said not possible for the NOg guy to win the </s><s>[INST] bears is a further discussion. What was to you? The biggest difference between mixed martial arts and uh jujitsu. What are some interesting differences, some interesting insights even just about the grappling with in both sports. So [/INST] the biggest difference for me between Emma and jujitsu is first is the speed. Like jujitsu, you know, like a 10 minutes match. I can take my time. There's no dangers that that forces me to move fast. M. M. A. You have to be 100% sharp in fast from 1st 2nd of the fight because punches are coming, can get knocked out anytime one mistake you out, you just you don't have that. Like it's I don't have to worry about quick submission because it's all about the way my body is positioned. You know, my grips, it's easy to avoid, it's easier to see it coming. Like it's like a quick submission or surprise. It only works if you make a mistake, if you're not correct positioning otherwise it's impossible. It's extremely difficult. Mm. A It's not I mean one split second mistake and when the person comes you have to respond, you have to match his space. I mean you can't slow down but it's like you you're forced to respond so much fast. It's it's a it's a lot more physical, a lot more and you need to be physical, much better conditioning faster. Its explosiveness, it's much harder. </s><s>[INST] Is it possible in M. m. A. two calm things down? [/INST] They change the rules. Yeah, five minute rules. Now, </s><s>[INST] I see. So like uh I just meant actually technically speaking is there ways to to take an opponent that's being exceptionally aggressive? You [/INST] can clinch, but then he takes you down, he keeps moving something is hard to control that pace. You can you don't if you play defense, you you save more energy than if you try to be the aggressor and respond. And </s><s>[INST] even getting to the clinch is very difficult [/INST] and you have no way to hold yourself there. So that was the biggest challenge for me in M. M. A. Is the speed because I'm a very slow start fighter if you look at my matches, I start very slow because if I go hard? You know, I fatigue faster. So for me, that was the hardest part is to start fast. What </s><s>[INST] about on the ground? Is there something different more challenging on the ground? [/INST] Being in the bottom? Yes, this punches punch coming. </s><s>[INST] How fundamentally different is you just punches on the [/INST] ground changes everything, everything which </s><s>[INST] which part's [/INST] the distance that you allow your opponent to be on you. The techniques that you choose to apply. You know, you you have to your body has to be aware of the punches and, you know, you are a lot more limited on your attacks. </s><s>[INST] So, you're you're known for your clothes guards. How does your close guard have to adjust? How does the position of your hands have to adjust when you're on the bottom of closed guard? [/INST] So in the garden, especially the close guard, you have to keep the person very close to you or you have to kick him away. That's in the guard. Is either I'm hugging you or get away from </s><s>[INST] energy jesse, you're allowed to have a middle. [/INST] Yeah, there's a lot you can allow the person to be. </s><s>[INST] What about getting uh arm lock or triangle submissions from the from the guard. Is that fundamentally different? Because you don't have the middle [/INST] game? It's much harder. There's barely no open guard in a meme a very little because the open guard, there's a distance between you and him. There's a distance to how you you cannot control is much harder to control that punch coming. And I have to position myself away to block that. And it limited my attacks, my options of attacks. </s><s>[INST] Is there a reason, correct me if I'm wrong, but I don't think you do open guard much in jujitsu and NOg. Is there a reason for that? [/INST] It's harder with the explosive uh explosive person because it's not when they're moving fast then you have to try to slow them down. </s><s>[INST] So you like guards that allow you to control the [/INST] person </s><s>[INST] and closed guard is the ultimate [/INST] control. </s><s>[INST] It's [/INST] not ultimate control but close guard put in the position that I'm attacking, you're defending. You cannot attack me from my close guard. We can argue that this might be one or two attacks but it's like very very very limited and depends where you're fighting against. </s><s>[INST] I hate the close guard being on top well against a good close guarded [/INST] everyone likes. It's terrible, </s><s>[INST] it's horrible. [/INST] It's one sided. So even the garden is one sided. The person the border has the advantage. I can be completely relaxed in my close guard and I cannot be completely relaxed. You </s>

<s>[INST] of. So that's that's a great thing at a company that works on uh write me like celebrating humanity in general broadly speaking and that's a great thing to have in a company that works on ai because that's not often the thing that's celebrated in ai companies often just raw great engineering. Um just celebrating the humanity. That's great especially from a leadership position. Um Well what do you think about the movie her? Let me ask you that before I talk before I talk to you about because it's not effective to is and was not just about emotion. So I'd love to talk to you about smart. I but before that let me just jump into the movie her do you think we'll have a deep meaningful connection with increasingly deep and meaningful connections with computers. Is that a compelling thing to you? Something [/INST] that's already happening. The thing, I love them. I love the movie her by the way. But the thing I love the most about this movie is it demonstrates how technology can be a conduit for positive behavior change. So I forgot the guy's name in the movie </s><s>[INST] whatever. [/INST] So Theodore was like really depressed right? And he just didn't want to get out of bed and he just he was just like done with with life, right? And Samantha right? </s><s>[INST] Samantha, [/INST] she just knew him so well. She had she was emotionally intelligent and so she could persuade him and motivate him to change his behavior and she got him out and they went to the beach together. And I think that represents the promise of emotion. Ai if done well this technology can help us live happier lives more productive lives, healthier lives more connected lives. So um that's the part that I love about the movie. Obviously it's Hollywood so it takes a twist and whatever. But but the but the key notion that technology with emotion ai can persuade you to be a better version of who you are. I think that's awesome. </s><s>[INST] Well what about the twist? You don't [/INST] you </s><s>[INST] don't think it's good for spoiler alert that Samantha starts feeling a bit of a distance and basically leaves Theodore. Um You don't think that's a good feature. That's a that you think that's a bug or a feature? [/INST] Well I think what went wrong is Theodore became really attached to Samantha. Like I think he kind of fell in love </s><s>[INST] with. [/INST] I mean I think that </s><s>[INST] I think she was putting out the signal is an intimate relationship right? There's a deep intimacy to it. [/INST] Right. But what does what does what does that mean? What does that </s><s>[INST] mean? [/INST] Right? What does that mean? Right. Yeah. Well when </s><s>[INST] he realized which is such a human thing of jealousy when you realized that Samantha was talking to like thousands of people. [/INST] Yeah. Did not go well, you </s><s>[INST] know that doesn't from a computer perspective, like that doesn't take Anything away from what we have. It's like you're getting jealous of Windows 98 for being used by millions of people. But [/INST] it's like it's like not liking the Alexa. </s><s>[INST] Yeah, it's [/INST] a bunch of, you know other families, </s><s>[INST] but I think Alexa currently is just a servant. It tells you about the weather. It's not doesn't do the intimate deep connection. And I think there is something really powerful about that, the intimacy of a connection with an Ai system that would have to respect and play the human game of of jealousy, of love of heartbreak and all that kind of stuff, which no matter, does seem to be pretty good at. I think she, this ai systems knows what it's doing. [/INST] Well actually, let me ask you this, but I don't think </s><s>[INST] she was talking to anyone else. [/INST] You don't think so. You think she was just done with Theodore? </s><s>[INST] Yeah, she knew that. Yeah. And then she wanted to really [/INST] move on. </s><s>[INST] She didn't have the guts to just break it off cleanly. [/INST] Okay. </s><s>[INST] She wanted to put in the no, I don't know. [/INST] She could have ghosted him, </s><s>[INST] She could have [/INST] I'm </s><s>[INST] sorry there's our engineers. [/INST] Oh God. </s><s>[INST] But I think those are really I I honestly think some of that some of it is Hollywood, but some of that is features from an engineering perspective, Not, not not a bug. I think, I think ai systems that can leave us now, this is for more social robotics than it is for anything that's useful. Like I hated Wikipedia said, you know, I need a break right now, like no, no, I need you. Um but if it's just purely uh for for companionship, then I think the ability to leave is really powerful. I never [/INST] thought of that. So that's so, so fascinating because I've always taken the human perspective, right? Um like for example, we had a at home, right? And my son loved it and then the company ran out of money and so they had to basically shut down like Jimbo basically died, right? And it was so interesting to me because we have a lot of gadgets at home and a lot of them break and my son never cares about it, right? Like if far Alexa stopped working tomorrow, I don't think he really care. But when Jimbo stopped working, it was traumatic, like he got really upset and as a parent that like made me think about this deeply, right? Did I was I comfortable with that? I liked the connection they had because I think it was a positive relationship, but I was surprised that it affected him emotionally so much and I think there's a broader question here, right? As we build socially and emotionally intelligent machines, what does that mean about our relationship with them? And then we're broadly our relationship with one another, right? Because this machine is going to be programmed to be amazing and empathy by definition, right? It's gonna always be there for you, it's not gonna get bored. Um In fact there's a chatbot in china, Shall shall I um and it's like the number two or three most popular app and it basically is just a confidante and you can tell it anything you want and people use it for all sorts of things they confide and like um domestic violence or societal attempts or you know if they have challenges at work, I don't know what that I don't know if I'm coming, I don't know how I feel about that, I think about that a lot. </s>

<s>[INST] Yeah, so it's funny, I'm very much with you on the 5 to 10 years, maybe 10 years, but you made it, I'm not sure how you made a sound, but for some people that seems that might seem like really far away and then for other people, uh it might seem like very close, there's a lot of fundamental questions about how much game theories isn't this whole thing, so like how much is this simply uh collision avoidance problem and how much of it is uh you're still interacting with other humans in the scene and you're trying to create an experience that's compelling. Do you want to get from point A to point B quickly? You want to navigate the scene in a safe way? But you also want to show some level of aggression because well, certainly this is why you're screwed in India because you have [/INST] to show a great Jersey. Uh </s><s>[INST] so like uh or new york or basically any major city, but I think it's probably Ellen that I talked the most about this, which is uh surprised the level of which they're not considering human beings as a huge problem in this as a source of problem. Like uh driving is fundamentally um, a robot on robot versus the environment problem versus like, you know, you can just consider humans not part of the problem. I used to think humans are almost certainly have to be uh modeled really well pedestrians and cyclists and humans inside of the cars. You have to have like mental models for them. You cannot just see it as objects, but more and more. It's like the, it's the same kind of intuition breaking thing that self supervised learning does, which is, well, maybe through the learning you'll get all the human like human information you need, right? Like maybe you'll get it just with enough data, you don't need to have explicit good models of human behavior. Maybe you get her through the data. So I mean, my skepticism also just knowing a lot of automotive companies and how difficult it is to be innovative. I was skeptical that they would be able at scale to convert the driving seen across the world into digital form such that you can create this data engine at scale. And the fact that Tesla is at least getting there or are already there makes me think that it's now starting to be coupled to the self supervised, um, learning vision, which is like, if that's gonna work if through purely this process you can get really far, then maybe you can solve driving that way. I don't know. I, I tend to believe we we don't give enough credit to the, to the how amazing humans are both at driving and at supervising autonomous systems and also we don't, this is, I wish we were, I wish there was much more drivers sensing inside Tesla's and much deeper consideration of human factors like understanding psychology and drowsiness and all those kinds of things when the car does more and more of the work, how to uh, keep utilizing the little human supervision that I needed to keep this whole thing safe. I mean, it's a fascinating dance of human robot interaction. To me, autonomous driving for a long time Is a human robot interaction problem. It is not a robotics problem or computer vision problem. Like you have to have a human in the loop. But so which is why I think it's 10 years plus, but I do think there will be a bunch of cities and contexts where you know uh you're restricted, it will work really, really damn well. [/INST] So I think for me that gets five if I'm being optimistic and it's going to be five for a lot of cases and then plus I agree with you 10 plus basically uh if you want to recover most of the contiguous United States or something </s><s>[INST] Interesting. So my optimistic is five and pessimistic is 30. I have a long tail on this one. I've watched enough driving videos, I've watched enough pedestrians to think like we may be like there's a small part of me still not a small like a pretty big part of me that thinks we will have to build a G. I. To solve driving. [/INST] Oh well </s>

<s>[INST] So what blows your mind most about that visualization [/INST] when you just sit back and think about it? It's a boring book. So 950 pages, 95% of the book. Hunter Gather is kind of doing their thing. I'm sure there's obviously some major cognitive and advancements along the way in language. And I'm sure the bone arrow comes around at some point. So tiny things, but it's like, oh, now we have 400 pages till the next thing. But then you get to page 950 and things start moving. </s><s>[INST] Recorded history starts at 976. Right, right. So basically the bottom row is when anything interesting happens. There's a bunch of agriculture for a while before we know anything about it. And then recorded history starts. [/INST] Yeah, right. Yeah, 25 pages of actual recorded history. So when we think of prehistoric, we're talking about pages one through 975 of the book. And then history is page 976 to 1000. If you were reading the book, it would be like epilogue, A.D., the last little 10 pages of the book. And we think of A.D. as super long, 2,000 years. The Roman Empire 2,000 years ago. Like that's so long. Human history has been going on for over 2,000 centuries. Like that is, it's hard to wrap your head around. And this is, I mean, even that's just the end of a very long road. Like the 100,000 years before that, it's not like that was that different. So it's just, there's been people like us that have emotions like us, that have physical sensations like us for so, so long. And who are they all? And what was their life like? And it's, I think we have no idea what it was like to be them. The thing that's craziest about the people of the far past is not just that they had different lives, they had different fears, they had different dangers and different responsibilities and they lived in tribes and everything, but they didn't know anything. Like, we just take it for granted that we're born on top of this Tower of Knowledge and from the very beginning, we know that the Earth is a ball floating in space and we know that we're going to die one day and we know that, we evolved from animals and all the, those were all like incredible epiphanies quite recently. And the people a long time ago, they just had no idea what was going on. And like, I'm kind of jealous because I feel like it might've been scary to not know what's going on, but it also I feel like would be, you'd have a sense of awe and wonder all the time and you don't know what's gonna happen next. And once you learn, you're kind of like, </s><s>[INST] oh, that's like, it's a little grim. But they probably had the same capacity for consciousness to experience the world, to wander about the world, maybe to construct narratives about the world and myths and so on. They just had less grounded systematic facts to play with. They still probably felt the narratives, [/INST] the myths that constructed as intensely as we do. Oh yeah, they also fell in love. They also had friends and they had falling outs with friends. They didn't shower much though. </s><s>[INST] They did not smell nice. Maybe they did. Maybe beauty's in the eye of the beholder. [/INST] Maybe it's all like relative. Yeah. How many people in history have experienced a hot shower? Like almost none. That's like, when were hot showers invented 100 years ago? Like less. So like George Washington never had a hot shower. It's like, it's just kind of weird. Like he took cold showers all the time or like. Yeah. And again, we just take this for granted, but that's like an unbelievable life experience to have a rain, a controlled little booth where it rains hot water on your head. And then you get out and it's not everywhere. It's like contained. That was like, you know, a lot of people probably lived and died with never experiencing hot water. Maybe they had a way to heat water over a fire, but like, then it's, I don't know. It's just like, there's so many things about our lives now </s><s>[INST] that are completely, just total anomaly. It makes me wonder like, what is the thing they would notice the most? I mean, the sewage system, like it doesn't smell in cities. Incredible. What does the sewer system do? I mean, it gets rid of waste efficiently, such that we don't have to confront it both with any of our senses. And that's probably wasn't there. I mean, what else? Plus all the medical stuff associated with sewage. [/INST] Yeah, I mean, how about the disease? Yeah. How about the cockroaches and the rats and the disease and the plagues and, you know. And then when they got, so they caught more diseases, but then when they caught the disease, they also didn't have treatment for it. So they often would die or they would just be a huge amount of pain. They also didn't know what the disease was. They didn't know about microbes. That was this new thing, the idea that these tiny little animals that are causing these diseases. So what do they think? In the bubonic plague, you know, in the black death, the 1300s people thought that it was an act of God because you know, God's angry at us. Cause why would, you know, why would you not think that if you didn't know what it was? And so the crazy thing is that these were the same primates. So I do know something about them. I know in some sense what it's like to be them, because I am a human as well, and to know that this particular primate, that I know what it's like to be, experienced such different things. And like this isn't, our life is not the life that this primate has experienced almost ever. </s>

<s>[INST] It's just basic fear. It's the same kind of fear that if you see, if you're a guy, see a cute girl at a bar, it's the fear associated with coming up to her. You like her asking her. It's like, what's the actual risk exactly? Um, [/INST] she'll say no thanks, I'm not. </s><s>[INST] And I guess the risk is like, that's going to be mentally difficult to deal with rejection. So just like it's mentally difficult to deal with failure if you had a bunch of ideas and you're excited about them and you implement them and you realize they're not good. That could be difficult to keep pushing through that. But I suppose that's life. You're supposed to, you know, perseverance through the failures and then the risk is low. So that's, and then the whole time through the fog up the mountain, you're looking for product market [/INST] fit. Yeah, that's right. So, you know, you have it when the usage of your product keeps growing without any marketing dollars or anything like that. It's just like more people keep coming back every week or month. So you kind of keep your basically watching your stats, nothing is working. You see these little wiggles of false hope in your metrics and you basically just keep talking to customers fixing the, improving the product, talk to customers, improve the product. Talk to customers with product, you know, and trying to run out of money, so be really scrappy. And then if you're lucky, you hit some kind of threshold, we're like, okay, the thing is good enough now or we hit on some use case and then it'll organically start to grow a bit and then then you have a whole different set of problems once you probably aren't fit, which is how do we scale this thing? How do we hire people? How do we, you know, hire an executive team or raise more money? And like, so the problem is totally changed, but um, </s><s>[INST] you were there through the whole thing. So that's the other question. That's fascinating. Again, back to the girl at the bar, how do you hire people? It's like how do you find good friends? How do you find good relationships and in this specific case, how do you hire good people, engineers, Executive. All of it. [/INST] One thing is, I've done a lot of reps on hiring at this point. So, um, Coinbase has about 5000 people. Um, probably the 1st 500 people or something, maybe in that range. Um you know, I interviewed every single one of those, but you have to remember there's probably like, I don't know, on average, maybe 10 people that we went in the process for everyone we hired or something. So it was like By the time that we had 500 employees, I had done like 5000 interviews or something. I was like very burned out in interviews. I had been doing some days. I did like seven interviews in a day or maybe um you know you you could do lots of interviews, maybe you wouldn't get burned out. But different interviews, </s><s>[INST] very different, very different, very different because you're uh so first of all, most of your interviews lead to rejection. Yeah. Which is also exhausting. [/INST] Yeah and there's a whole there's a whole part of the interview which is about candidate experience right? Sometimes you know it's not the right person but you want to make sure they have a good experience. Like if you're just exhausted and you're on your sixth interview and you're like well thanks for coming in and you wrap and you just and then like you're going to create a detractor. Someone who's out there like that company or brian was rude to me or whatever so I had to honestly I had to work on that a little bit in the early days. I was doing so many interviews like I needed to make sure that um when people came in I was like you know made them feel comfortable um ask them a couple of like warm up questions, just like how was it getting in the office? Like did you find it? Okay and like what you've been up to this week and not just like you know like a factory assembly line like boom boom boom like. Yeah but </s><s>[INST] also there's a moment because I've interviewed a bunch of people for like teams and stuff. There's also a moment when you early on, no it's not going to be a good fit and you still have to land that plane and and all that kind of stuff and that could get really really, really exhausting. So yeah. Anyway, sorry? So [/INST] Yeah so basically we tried, we tried so many things over the years to make interviews more efficient because it's a huge time sink for um the team. So you know we we basically will usually get them down to like 25 minutes. Um I've seen if you're trying to hire like a big team, let's say um you know of people who are like contractors or something, not necessarily full time employees. I've seen people actually do 10 minute interviews. Um You can even interview like 1000 people almost like in a week or something. I'm not sure if that quite works out but let me a little less than that but you can basically get six and six down an hour if you just need to get a team of 30 contractors for whatever purpose but if you're talking about full time employees, I usually do like 25 minutes. You know you're you're often times like 11 thing we've done is we put like a, um, like a google form online and it's like put some basic hurdles in there, like, you know, ask them um, to put in an answer which you can check in a spreadsheet if it was correct or not. And like there was some funny examples in the early days of coin base where we put in like brain teasers and stuff, but we don't do that anymore. We do like normal interviews. We do references the kinds of things I ask in interviews. Um, you know, it's usually like, I like to think about what, what do we need this person to accomplish in this role, right? And get really specific about that? It's like usually something pretty hard and then I'll ask them a question. It's like, tell me about a time you did X and um, or tell me about the hardest, the hardest kind of problem you had to solve in why and what did you do specifically to overcome it? Right? So I'm asking to see if they can actually do the stuff we need to get done. But then I'm also kind of asking like cultural questions if I'm interviewing for that. And so I'm trying to see like are they concise communicators, can they just give me a clear answer and stop talking. Some people like ramble on for like 5, 10 minutes if you ask the first question, some people are, you know, they're interrupter like church of interruption. So like they won't stop talking until you interrupt them. Which for me I'm always patient and wait. So that's weird. I'm looking to see for humility to like you know I'll tell you ask people tell me about a time something went really wrong like you had conflict or someone on the team or um what I'm kind of looking for is were they part of the solution or are they still holding on to like blame and criticism about that and be like well I told them they shouldn't do that way but they didn't listen to me. You know that these are all like bad signs so I'm looking for. Yeah can they get the job done? Well they work together on a team? Can they communicate effectively? Do they fit into our cultural values and you know. </s>

<s>[INST] it calms you and focuses you. What does it different. [/INST] I think it gets my vocal cords warmed up and it also do </s><s>[INST] you sing or speak that [/INST] I often sing them uh and unfortunately nobody hears and as I do this I start to evaluate whether or not I'm straining to get the words out or whether or not I'm straining to make them slow enough so that I can articulate them. So there are days when I have so much energy that I trying to speak faster than I should in order to articulate properly. There are other days when I'm tired and I can't sort of keep up with my thoughts and so what I try and do is assess that and then adjust the transmission, the rpm, so to speak. For instance, I can speak very quickly and then I can slow down so I can change the cadence of my voice and when you teach in the classroom you learn as you know because you're an excellent teacher. I've watched your lectures in the classroom as you teach in the classroom. When you want to slow down, every teacher knows, you turn to the whiteboard or chalkboard and you start writing right, it gives you a break and then you turn around and you fire back the kind of machine gun fire of of information and then you slow down or you underline something when you podcast, you don't have that opportunity right? There are no visuals in my podcast. So what I try and do is always get my voice warmed up and make sure that I'm thinking and speaking at approximately the same rate. And then I also do this thing and I put my vision into panoramic vision when I walk, which is very calming and then I actually start to remind myself of the purpose of podcasting. This sounds very mission statement. But you asked what I do. I remind myself first and foremost that what I want to communicate, what I want to come through is the beauty and utility of biology. And I only feel comfortable saying the word beauty publicly now about science things thanks to you because I think </s><s>[INST] love and beauty [/INST] love and love and beauty, but also darkness and hatred. And if you're talking about the lex Fridman podcast, you have to adjust that you have to address the shadow also the shadow side. But I think about the I want to communicate the the beauty and utility of biology and then I check my, my emotional state. I want to make sure that I'm not angry about anything and certainly if I am that I'm going to set it aside for the podcast because that's not a place for for my whatever I might be dealing with. I also really start to feel into the parts of the research and the papers I found that I really love because that's the part of me that I like the most frankly and on the podcast. If there's a paper like for instance, we have a paper, excuse me a podcast coming out soon about um heat as a tool, you know, sauna but some other things and in researching this, I learned so much about um these heat shock proteins and the use of sauna in Finland for increasing growth hormone, but also for the treatment of mental illness. And I realized I fell in love with this literature. It's just a beautiful literature. These people are true pioneers for doing this work. Now everyone's in the sauna, but this was 20 years ago. The way the experiments were done were amazing with all these finished people with thermal couples up their rectum to measure temperature swimming in pools. It's hilarious and great. And so I start to think about it. I think, you know, I just start to really access my love of the work and then when we finally sit down meaning my producer rob and I and record, I just sort of want to just bask in in sharing it. Just like the little version of me when I was six or seven, I used to spend all weekend reading the encyclopedia Guinness book of World Records, making my mother driving me places to introduce me to have this obsession with trapping animals when I was a kid, meet these people. And then on monday I would insist on giving a lecture in class. It was a little kid. So that's basically what it is. I just try and access that that childlike energy. And um so I want to be clear. The goal is always to make the information interesting, clear and actionable. And if it's also surprising then there then that's a bonus. But that's basically the process. But yeah, I'm I'm singing and talking and and getting into state and I used to feel very sheepish about sharing any of this is the first time I've ever shared it out loud. But but Rick was the one who encouraged me to find a process that works and continue to develop that process and not let anything get near that process people in my personal life. No, this and when it's time, it's like, I don't care what else is going on on. I'm I'm moving into that brain state </s><s>[INST] and that's probably a process like that for anything that you do in life that you take seriously. So the people who have perfected this is athletes like olympic level athletes, they have to have a process like this. You [/INST] know what? I think Tiger Woods actually was taught self hypnosis quite young. Um and use self hypnosis often during his tournaments, sometimes too great uh success and other times less. So </s><s>[INST] is there other places in life that you use kind of a protocol, like a mental protocol to get ready? [/INST] Many of the best areas of life are their own form of hypnosis, Right? Um you know that you're in hypnosis if for instance you're in a movie and something happens and you feel the emotional lift with with without being self conscious about it. Um Yes, I think that um one thing that we've tried to do in our houses around meal times to try and set a state that food isn't just something that we just throw down our throats and I'm fortunate that, you know, my partner cooks really well and so I try and give her the space to do that and that's the whole thing of her getting into state and then </s>

<s>[INST] but there is also a sacrifice of ego. That's a pretty powerful thing [/INST] from Beautiful. Would you </s><s>[INST] do sort of philosophically to ask you the question, would you do all the same things you're doing now if your name wasn't attached to it? Sort of if if you had to sacrifice the [/INST] ego put </s><s>[INST] another way or is your ego deeply tied in the decisions you've been making? [/INST] I hope not. I mean, I I believe I would certainly attempt to do the things without my name having to be attached with it, but um it's hard to do that in a corporation. Um legally that's the issue. Um if I were to do more open source things than Absolutely, like I don't I don't need my particular identity, my real identity associated with it. But um I think, you know, the appreciation that comes from doing something good and being able to see it and see people use it is pretty overwhelming and powerful. Um more so than maybe seeing your name in the headlines. </s><s>[INST] Let's talk about artificial intelligence a little bit if we could 70 years ago Alan Turing formulated the touring tests. To me, natural language is one of the most interesting spaces of uh problems that are tackled by artificial intelligence is the canonical problem of what it means to be intelligent. He formulated the turing test, let me ask sort of the broad question. How hard do you think is it too pass the turing test in the space of language. [/INST] Just from a very practical standpoint I think where we are now and and for At least years out is one where the artificial intelligence machine learning deep learning models can bubble up, interesting this very, very quickly and pair that with um human discretion around severity, around depth, around nuance and and meaning. I think for me um the chasm across for general intelligences to be able to explain why and the meaning behind something behind </s><s>[INST] the decision. [/INST] So being behind the </s><s>[INST] decision [/INST] to um such sets of data. So the </s><s>[INST] explain ability part is kind of essential to be able to explain using natural language why the decisions were made, that kind of thing. [/INST] Yeah, I mean I think that's one of our biggest risk and artificial intelligence going forward is we are building a lot of black boxes that can't necessarily explain why they made a decision or what criteria they used to make the decision and we're trusting them more and more from lending decisions to content recommendation to driving to health. Um like you know a lot of us have watches that tell us to understand I was deciding that. I mean that one's pretty, pretty simple but you can imagine how complex they get and </s><s>[INST] being able to explain the reasoning behind some of those recommendations seems to be an essential part, although it was [/INST] a very hard problem because sometimes even we can't explain why we make that. That's </s><s>[INST] what I was. I think we're being a sometimes a little bit unfair for two artificial intelligence systems because we're not very good at these some of these things. Do you think? I apologize for the ridiculous, romanticized question, but on that line of thought, do you think we'll ever be able to build a system like in the movie her that you could fall in love with? So have that kind of deep connection with [/INST] hasn't that already happened, hasn't someone in Japan falling in love with his ai </s><s>[INST] there's always going to be somebody that does that kind of thing. I mean a much larger scale of actually building relationships of being deeper connections. It doesn't have to be loved, but it just deeper connections with artificial intelligence systems. So you mentioned explained it was [/INST] a loss of function of the artificial intelligence and more a function of the individual and how they find meaning and where they find meaning. </s><s>[INST] Do you think we humans can find meaning in technology in this kind of [/INST] What? 100% and I don't necessarily think it's a negative. Um but I you know, it's constantly going to evolve. So I don't know but meaning is something that's entirely subjective and I don't think it's going to be a function of finding the magic algorithm that enables everyone to love it. Um but maybe </s><s>[INST] but that question really gets at the difference between human and machine that you had a little bit of an exchange with Elon musk basically. I mean it's a trivial version of that, but I think there's a more fundamental question of is it possible to tell the difference between bought in a human and do you think it's If we look into the future, 10, 20 years out, do you think it would be possible? Or is it even necessary to tell the difference in the digital space between a human and a robot? Can we have fulfilling relationships with each or do we need to tell the difference between them? [/INST] I think it's certainly useful in certain problem domains to be able to tell the difference. Um I think in others it might not be that's useful and anything is </s><s>[INST] possible for us today to tell the difference. It's the reverse the meta of the turing test. [/INST] Well, what's interesting is, um I think the technology to create is moving much faster than the technology to detect. </s><s>[INST] You think so. So if you look at like adversarial machine learning, there's a lot of systems that try to fool machine learning systems. And at least for me, the hope is that the technology to defend will always be right there. At least your senses that [/INST] um I don't know if they'll be right there. I mean it's it's a race, right? So the detection technologies have to be Two or 10 steps ahead of the creation technologies. This is a problem that I think the financial industry will face more and more because a lot of our risk models, for instance, are built around identity payments ultimately comes down to identity. And you can imagine a world where um all this conversation around Deepfakes goes towards the direction of driver's license or passports or state identities and people construct identities in order to get through a system such as ours to start accepting credit cards or into the cash up. Um, and those technologies seem to be moving very very quickly. Our ability to detect them I think is probably lagging at this point, but um, certainly with more focus, we can get ahead of it. Um, but this is going to touch everything. So, um, I I think it's it's like security and we're never going to be able to build a perfect detection system. We're only going to be able to, you know, what we should be focused on is is the speed of evolving it and being able to take signals that show correctness or errors as quickly as possible and move um, and to be able to build that into our newer models or the or the self learning models. </s>

<s>[INST] But you wouldn't pay 10 times like um the problem is there's not financial, like it doesn't make sense to, to, to uh revamp the infrastructure to make land keeping easier. It does make sense to reinvent the infrastructure, if you have a large fleet of autonomous vehicles now you change what it means to own cars, you change the nature of transportation, I mean, but that for that, you need autonomous vehicles. Let me ask you about Waymo, then I've gotten a bunch of chances to, to ride in in a way mo um self driving car and there, I don't know if you call them self driving, but [/INST] well, I mean, I, I wrote in one before they were called Waymo still at X. </s><s>[INST] So there's currently, which was a big another surprisingly if I didn't think it would happen, which is, they have no driver [/INST] currently. Yeah, Chandler </s><s>[INST] in Chandler Arizona and I think they're thinking of doing that in Austin as well, but they're they're expanding, [/INST] although although, you know, and I do an annual check up on this. So, as of late last year, they were aiming for hundreds of rides a week, not thousands. And um, there is no one in the car, but certainly uh, safety people in the loop and it's not clear how many, you know what the ratio of cars to safety people is. </s><s>[INST] It wasn't obviously they're not 100% transparent about this, but [/INST] none of them. 100% this is a problem, but </s><s>[INST] at least the way there, I don't want to make definately, but they're saying there's no telling operation. Um, like there, I mean, okay, [/INST] and, and, and, and, and that sort of fits with, with Youtube videos. I've seen people being trapped in the car by a red cone on the, on the street and they do, they do have rescue vehicles that come and then a person gets in and drives it. Yeah, </s><s>[INST] But isn't it incredible to you? It wasn't to me to get in a car with no driver and watch the steering will turn like for somebody who has been studying, at least, certainly the human side of autonomous vehicles for many years and you've been doing it for way longer. Like, it was incredible to me that this was actually could happen. I don't care if that scale is 100 cars, This is not a demo, this is not, this is me as a [/INST] regular, the argument I have is that people make interpolations from that interpretations that here it's done. Um you know, it's just you know, we've solved it. No, we haven't yet. And that's my argument. Okay, </s><s>[INST] so I'd like to go to you, you keep a list of predictions on your amazing blog posts would be fun to go through them. But before that, let me ask you about this. You have uh you have a harshness to you sometimes in your criticisms of what is [/INST] chrissy uh pipe. </s><s>[INST] Uh so like because people extrapolate like you said, and they kind of buy into the hype and then they they kind of start to think that the technology is way better than it is. But let me ask, you may be a difficult question. So do you think if you look at history of progress, don't you think to achieve the quote impossible? You have to believe that is possible. [/INST] Absolutely. Yeah. Look, his his his his his two Great runs great. Unbelievable, 19 oh three. First human um how human uh heavier than air flight? Yeah, 1969. We land on the moon, 66 years. I'm 66 years old in my lifetime. That span of my lifetime barely, you know, flying. I don't know what it was. 50 ft or the length of the first flight or something to landing on the moon. Unbelievable. </s><s>[INST] Fantastic. But that requires by the way, one of the wright brothers, both of them, but one of them didn't believe it's impossible, like a year before, right? So like not just possible soon, but [/INST] like, so you know, how </s><s>[INST] important is it to believe and be optimistic is what I guess. [/INST] So yeah, it is important. It's when it goes crazy when when when, you know you said, what was the word used? My bad harshness, harshness? Yes. I just get so frustrated. Yes. When when people make these leaps and tell me that I'm that I don't understand. I, you know, Yeah, because Just from I robot, which I was co founder of. Yeah, I don't know the exact numbers now because I haven't it's 10 years since I stepped off the board, but I believe it's well over 30 million robots cleaning houses from that one company and now there's lots of other companies yesterday. Um was that a crazy idea that we had to believe uh In 2002 when we released it? Yeah, that was we had we had, you know, I believe that it could be done. Let </s><s>[INST] me ask you about this. So I robot, one of the greatest robotics companies ever in terms of manufacturing, creating a robot that actually works in the real world is probably the greatest robotics company ever. You were the co founder of it. Um If if the Rodney brooks of today talk to the Rodney of back then, what would you tell them? Because I have a sense that would you put him on the back and say what you're doing is going to fail, but go at it anyway, that's what I'm referring to with the harshness. You've accomplished an incredible thing. They're one of several things we'll talk about. Well, like that's what I'm trying to get at that line. [/INST] No, it's, it's when my harshness is reserved for people who are not doing it, who claim it's just, well, this shows that it's just gonna happen. </s>

<s>[INST] and slow down. And on top of all this, you know, we've been talking about astronomy and cosmology. There's been a lot of exciting breakthroughs in detection and exploration of black holes. So the boomeranging photons that you're talking about, there's been a lot of work on photon rings and just all the fun stuff going on outside the black holes. So all the garbage outside is actually might be the thing that holds a key to understanding what's going on inside. And there's the Hawking radiation. There's all kinds of fascinating stuff like, I mean, there's trippy stuff about black holes that I can't even, most people don't understand. I mean, the holographic principle with the plate and the information being stored potentially outside of the black hole, I don't even, I can't even comprehend how you can project a three-dimensional object onto a 2D and some hostile information where it doesn't destroy it. And if it does destroy it, challenging all the physics. All of this is very interesting, especially for kind of more practical applications of how the black hole can be used for propulsion. [/INST] Yeah. I mean, it may be that black holes are used in all sorts of ways by advanced civilizations. I think, again, it's been a popular idea in science fiction or science fiction trope that Sagittarius A star, the supermassive black hole in the center of our galaxy could be the best place to look for intelligent life in the universe because it is a giant engine in a way. You know, a unique capability of a black hole is you can basically throw matter into it and you can get these jets that come out, the accretion disks and the jets that fly out. And so you can more or less use them to convert matter into energy V equals MC squared. And there's pretty much nothing else except for, you know, annihilation with its own antiparticle as a way of doing that. So to have some unique properties, you could perhaps power civilization by just throwing garbage into a black hole, right? Just throwing asteroids in and power your civilization with as much energy as you really would ever plausibly need. And you could also use them to accelerate the way across the universe. And you can even imagine using small artificial black holes as thermal generators, right? So the Hawking radiation from them kind of exponentially increases as they get smaller and smaller in size. And so a very small black hole, one you could almost imagine like holding in your hand would be a fairly significant heat source. And so that raises all sorts of prospects about how you might use that in an engineering context to power your </s><s>[INST] civilization as well. You have a video on becoming a Kardashev type one civilization. What's our [/INST] hope for doing that? We're a few orders of magnitude away from that. Yeah, it is surprising. I think people tend to think that we're close to this scale. The Kardashev type one is defined as a civilization, which is using as much energy as is essentially incident upon the planet from the star. So that's a border, I think, for the Earth of something like 10 to the five terrores or 10 to the seven terrores is a gigantic amount of energy. And we're using a tiny, tiny, tiny fraction of that right now. So if you became a Kardashev type one civilization, which is seen not necessarily as a goal into itself, I think people think, well, why are we aspiring to become this energy hungry civilization? Surely our energy needs might improve our efficiency or something as time goes on. But ultimately, the more energy you have access to, the greater your capabilities will be. I mean, if you want to lift Mount Everest into space, there is just a calculable amount of potential energy change that that's going to take in order to accomplish that. And the more energy you have access to as a civilization, then clearly the easier that energy achievement is going to be. So it depends on what your aspirations are as a civilization. It might not be something you want to ever do, but- </s><s>[INST] Well, but we should make clear that lifting heavy things isn't the only thing, it's just doing work. So it could be computation. It could be more and more and more and more sophisticated and larger and larger and larger computation, which is it does seem where we're headed with the very fast increase in the scale and the quality of our computation outside the human brain, artificial computation. [/INST] Yeah, yeah, yeah. I mean, computation is a great example of, I mean, already I think some like 10% of US power electricity use is going towards the supercomputing centers. So there's a vast amount of our current energy needs, which are already going towards computing will surely only increase or only increase over time. If we start ever doing anything like mind uploading or creating simulated realities, that cost will surely become almost a dominant source of our energy requirements at that point if civilization completely moves over to this post-humanism stage. And so it's not unreasonable that our energy needs would continue to grow, certainly historically, they always have at about 2% per year. And so if that continues, there is going to be a certain point where you're running up against the amount of energy, which you can harvest because you're using every, even if you cover the entire plant and solar panels, there's no more energy to be had. And so this is a, you know, there's a few ways of achieving this. I sort of talked about in the video how there were several renewable energy sources that we're excited about, like geothermal, wind power waves, but pretty much all of those don't really scratch the surface or don't really scratch the itch of going to a Kardashev type one civilization. They're meaningful now. I would never tell anybody don't do wind power now, because it's clearly used for our current stage of civilization, but it's not, it's going to be a pretty negligible fraction of our energy requirements if we got to that stage of development. And so there has to be a breakthrough in either our ability to harvest solar energy, which would require maybe something like a space array of solar panels of beaming the energy back down or some developments and innovations in nuclear fusion that would allow us to essentially reproduce the same process of what's producing the solar photons, but here on Earth. But even that comes with some consequences. If you're generating the energy here on Earth and you're doing work on it on Earth, then that work is going to produce waste heat and that waste heat is going to increase the ambient temperature of the planet. And so whether, even if this isn't really a greenhouse effect that you're increasing the temperature of the planet. This is just the amount of computers that are churning. You put your hand to a computer, you can feel the warmth coming off them. If you do that much work of literally the entire instant energy of the planet is doing that work, the planet is going to warm up significantly as a result of that. And so that clearly indicates that this is not a sustainable path, that civilizations as they approach Kardashev Type 1 are gonna have to leave planet earth, which is really the point of that video to show that it's a Kardashev type 1 civilization, even though it's defined as instant energy upon a planet that is not a species that is going to still be living on their planet, at least in isolation. They will have to be harvesting energy from afar. They will have to be you doing work on that energy outside of their planet because, otherwise you're going to that is gonna still be living on their planet, at least, in isolation. They will have to be harvesting energy from afar, they will have to be doing work on that energy outside of their planet because, otherwise, you're gonna dramatically change the environment in which you live. </s>

<s>[INST] I mean there's, there's a pragmatic aspect with [/INST] the fact that complex numbers were there. I love that the fact that I could write and the arrays constructs and that reduction was there very simple to write summations and, and broadcasting was there, I could do addition of whole arrays. Um, so that was cool. Those are some things I loved about it. </s><s>[INST] I don't know what to start talking to you about because you've been you've created so many incredible projects that basically changed the whole landscape of programming. But okay, let's start with, let's go chronologically with cyp I, you create a site pi over two decades ago now. Yeah, I love [/INST] to talk about. It was really my baby. </s><s>[INST] What is it, what was its goal? What is its goal? How does it work? [/INST] Yeah. Fantastic. So cyp I was effectively here, I'm using python to do stuff that I previously used to have to use. And I was using numeric, which is an array library that made a lot of it possible. But you know, there's things that are missing, like I didn't have an ordinary differential equation solver, I could just call, right, I didn't have integration. Yeah, I wanted to integrate this function. Okay, well I don't have just a function, I can call to do that. Um these are things I remember being critical things that I was missing optimization. I just want to pass a function of an optimizer and have it tell me what the optimal value is. Uh those things like, well why don't we just write a library that adds these tools and I started posting the main list and their previously been, you know, people have discussed, remember Conrad Hanson saying, wouldn't it be great if we had this optimizer library or David Ash would say this stuff and I'm, you know, I'm a ambitious ambitions the wrong word and eager and uh, probably more time than since I was, you know, poor graduate student. My wife thinks I'm working on my PhD and I am, but part of a PhD that I loved was the fact that it's exploratory. You're not just, you know, taking orders, fulfilling a list of things to do, you're trying to figure out what to do. And so I thought, well, you know, I'm writing tools for my own use and a PhD. So I'll just start this project. And so 99 98 was when I first started to write libraries for python, especially when I fell in love with python, 98. I said, well, there's just a few things missing. Like, oh, I need a reader to read dot com files. I was in medical imaging and dot com was a format that I want to load that in the python. Okay, how do I write a reader for that? So I wrote something called It was an I. O package. Right? And that was my very first extension module, which is c. So I wrote c code to extend python so that the positive python, I could write things more easily. That that combination kind of hooked me. It was the idea that I could, here's this powerful tool. I can use, the scripting language and a high level language to think about, but that I can extend easily easily in the N. C that easily for me because I knew enough C. And then ghetto had written a link. I mean the only at the hard part of extending python was something called the way memory management networks and you have to reference county and so there's there's a tracking of reference county you have to do manually and if you don't have to have memory leaks and so that's hard. Plus then see you know it's just much more you have to put more effort into it. It's not just have to now think pointers and have to think about stuff that is different. I have to kind of your like putting a new cartridge in your brain like okay I'm thinking about. Um right now I'm thinking about programming and their distinct modules you end up having to think about. So it's harder when I was just in python I can just think about M. R. I. And high level writing. Um but I could do that and that kind of I liked it. I found that to be enjoyable and fun and so I ended up oh let me just add a bunch of stuff to python to do integration. Well and the cool thing is that you know the power of the internet just looking around and I found others this net live which has hundreds of four chan routines that people have written In the 60s and the 70s and the 80's In 1477 Fortunately it wasn't 1416 have been boarded the FORTRAN 77 Unfortunately, is actually really great Language, fortunate 90 probably is my favorite fortune because it's also it's got complex numbers got a raise and it's pretty high level. Now the problem with it is you never want to write a program in 14 90 or 14 77 but it's totally fine to write a subroutine in. Right. And so and then fortunate kinda got a little off course when they try to compete with C. Plus plus. But at the time I just want libraries that do something like oh here's an order interest equation, here's integration. Here's runs cut integration Already done. I don't have to think about that algorithm and you could but it's nice to have somebody who already done one and tested it. And so I sort of started this journey in 98. Really look back at the mailing list. There's sort of this this productive era of me writing an extension module to connect Runge cut immigration to python and making an ordinary additional equation solver and then releasing that as a package called Deepak. I think I called it then quad pack. And then I just made these packages eventually that became multi pack because they originally modular. You can install them separately. But a massive problem in python was actually just getting your stuff installed at the time releasing software for me like today. It's people think what does that mean? Well then it meant some poorly written web page, I have some bad web page up and I put a tar ball, just a g zip tar ball of source code. That was the release. </s>

<s>[INST] So, you know, the natural question when people hear breakthrough with the, with the inertial confinement with the magnetic confinement is, so when will we have commercial reactors, power plants that are actually producing electricity? What's your sense looking out into the future? When do you think you can envision a future where we have actual electricity coming from nuclear fusion? [/INST] Partly driven by us, but in other places as well too. So there's the advent, what's, you know, what's so different now than three or four years ago, like we launched around four years ago, what's so different now is, is the advent of a very nascent, but seemingly robust, like commercial fusion endeavor. So it's not just Commonwealth fusion systems. There's something like 20 plus, you know, companies. There's a sector now. There's a sector. They actually, they actually have something called the fusion industry association, which is, if you viewers want to go see this, this describes the difference and they've got this plethora of approaches. Like I haven't even described all the approaches. I've basically described the main line approaches. approaches, you know, and they're all at varying degrees of technical and scientific maturity with very huge different, you know, balances between them. But what they share is that because they're going out and finding, getting funding from the private sector is that their stated goals are about getting fusion into place so that both it meets the investor's demands, which are interesting, right, and the timescales of that, but also it's like, well, there's going to, and why? It's because it's easy, there's going, there's this enormous push, driver about getting carbon free energy sources out into the market and whoever figures those out is going to be both very, it's going to be very important geopolitically, but also economically as well too. So it's a different kind of bat, I guess, or a different kind of gamble that you're taking with fusion, but it's so disruptive that it's like there's essentially a class of investors and teams that are ready to go after it as well too. So what do they share in this? They typically share getting after fusion on a time scale so that could it have any relevance towards climate change, battling climate change? And I would say this is difficult, but it's fairly easy because it's math. So what you do is you actually go to some target like 2050 or 2060, something like this, and say, I want to be blank percent of the world's market of electricity or something like that. And we know historically what it takes to evolve and distribute these kinds of technologies because every technology takes some period of time, it's so-called S-curve, it's basically, everything follows a logarithmic curve, exponential type curve, it's a straight line of log plot. And like you look at wind, solar, fission, they all follow the same thing. So it's easy. You take that curve and you go at slope and you work backwards and you go, if you don't start in the early 2030s, like it's not going to have a significant impact by that time. So all of them share this idea. And in fact, it's not just the companies now. The US federal government has a program that was started last year that said, we should be looking to try to get like the first, and what do I mean by like, what does it mean to start? That you've got something that's putting electricity on the grid, a pilot, what we call it. And if that can get started like in the early 2030s, the idea of ramping it up makes sense. It's math, right? So that's the ambition then the question is, and actually this is different because the government program, and they vary around in this. So for example, the United Kingdom's government idea was to get the first one on by 2040. And China has ambitions probably middle 2030s, or maybe a little bit later and Europe, you know, continental Europe is, it's a little bit, I'm not exactly sure where it is, but it's like later, it's like 2050 or 2060, because it's mostly linked to the eater timeline as well too. The fusion companies, which makes sense, it's like, of course they've got the most aggressive timelines. It's like, we're going to map the human genome faster as well too, right? So it's interesting about where we are. And I think, you know, my, we're not all the way there, but my intuition tells me we're probably going to have a couple of cracks at it actually on that timeline. So this is where we have to be careful though. You say commercial fusion, you know, what does that mean? Commercial fusion to me means that you're actually have a known quantity about what it costs, what it costs to build and what it costs to operate the reliability of putting energy on the grid. That's commercial fusion. It's like, so it turns out that that's not necessarily exactly the first fusion devices that put electricity on the grid, because you got to, there's a learning curve to get like better and better at it. </s>

<s>[INST] Buy Bitcoin and invest in the stock market with as little as $1. It's one of the best design interfaces of an app that I've ever used to me. Good design is when everything is easy and natural. Bad design is when the app gets in the way either because it's buggy or because it tries too hard to be helpful. I'm looking at you clip E for Microsoft even though I love you Anyway, there's a big part of my brain heart that loves to design things and also to appreciate great design by others. So again, if you get cash out from the app store, google play and use the code lex podcast, you get $10 cash up. Will also donate $10 to first an organization that is helping to advance robotics and stem education for young people around the world. And now here's my conversation with Richard Karp, You wrote that at the age of 13, you were first exposed to plane geometry and was wonder struck by the power and elegance of form of proofs. Are there problems, proves properties, ideas in plane geometry that from that time that you remember being mesmerized by or just enjoying to go through to prove various aspects. [/INST] So Michael Reagan told me this story about an experience he had when he was a young student. Mhm who was tossed out of his classroom for bad behavior and was wandering through the corridors of his school and came upon two older students Who were studying the problem of finding the shortest distance between two non overlapping circles. And Michael thought about it and said um You take the straight line between the two centers and the segment between the two circles is the shortest Because the straight line is the shortest distance between the two centres. In any other line connecting the circles would be on a longer line. And I thought, and he thought and I agreed that this was just elegance. The pure reasoning could come up with such a result, </s><s>[INST] certainly the shortest distance From the two Centers of the Circles as a straight line. Uh Could you once again say what's the next step in that proof? [/INST] Well, any any segment joining the two circles. Uh if you extend it by taking the radius on each side, You get a segment with a path with three edges Which connects the two Centers. And this has to be at least as long as the shortest path which is the straight line </s><s>[INST] straight line. Yeah, wow. Yeah, that is that's quite, quite simple. So what what is it about that elegance that you just find compelling? [/INST] Well, just that you could establish a effect about geometry beyond dispute by pure reasoning. Um I I also enjoy the challenge of solving puzzles in plane geometry. It was much more fun than the earlier mathematics courses which were mostly about arithmetic operations and manipulating them. </s><s>[INST] Was was there something about geometry itself? The slightly visual component of it [/INST] can vision? Absolutely. I lacked three dimensional vision. I wasn't very good at three dimensional vision. Being </s><s>[INST] Able to visualize three dimensional [/INST] Objects, three dimensional objects or surfaces, hyper planes and so on. Um So so there there, I didn't have an intuition but um For example the fact that the sum of the angles of a triangle is 180° it is proved convincingly. Um And it comes as a surprise that that can be done. </s><s>[INST] Why why is that surprising? The, well it is a surprising uh surprising idea I suppose. Uh Why is that proved difficult? [/INST] It's not that's the point, it's so easy and yet it's so convincing. </s><s>[INST] Do you remember what is the proof that it's uh as up to one of [/INST] you started a corner Helen and draw a line um parallel to the opposite side. And that line sort of try sects the angle Between the other two sides and uh you get a half plane which has two Add up to 180°. And it consists in the angles by the equality of alternate angles. What's it called? You get a correspondence between the angles created by this along the side of the triangle And the three angles of the triangle. </s><s>[INST] Has geometry had an impact on when you look into the future of your work with combinatorial algorithms. Has it had some kind of impact in terms of Yeah, being able to the puzzles, the visual aspects that were first. So compelling to you [/INST] not Euclidean geometry particularly. I think I use tools like linear programming and integer programming a lot, but those require high dimensional visualization. And so I tend to go by the algebraic properties. </s><s>[INST] Um Right, uh you go by the algebra, the linear algebra and not by the visualization. [/INST] Well, the interpretation in terms of mhm. For example, finding the highest point on Apollo veteran as in linear programming is motivating. But again, it I don't have the high dimensional intuition that would particularly inform me. So I sort of deep lean on the algebra. </s><s>[INST] So to linger on that point, what kind of visualization do you do you do when you're trying to think about? We'll get to combinatorial algorithms but just algorithms in general. What kind of what's inside your mind when you're thinking about designing algorithms or or even just tackling any any mathematical problem? [/INST] Well, I think that usually an algorithm is involves a repetition of some inner loop and and so I can sort of visualize the distance from the desired solution. Yeah, iterative, lee reducing until you finally hit the exact solution </s><s>[INST] and try to take steps that get you closer to the [/INST] right to take steps that get closer and having the certainty of converging. So it's that's right. It's basically the the mechanics of the algorithm is often very simple, but especially when you're trying something out on the computer. So for example, um I did some work on the traveling salesman problem and uh I could see there was a particular function that had to be minimized and it was fascinating to see the successive approaches to the minimum to the optimum. </s>

<s>[INST] So PDP what's the PDP seven? So it's a piece of hardware, you [/INST] know, it's a piece of part where it was one of the early machines made by Digital equipment, Corporation deck. Um and it was a mini computer, so called, it Had, I would have to look up the numbers exactly, but it had a very small amount of memory, maybe 16K-16 bit words or something like that, relatively slow. probably not super expensive. Maybe again, making this up, I'd have to look at $100,000 or something like that, </s><s>[INST] which is not super expensive in those days. It [/INST] was expensive, it was enough that you and I probably wouldn't be right by one, but a modest group of people could get together, but in any case in It came out, if I recall in 1964, so by 1969 it was getting a little obsolete and that's why it was little used </s><s>[INST] if you can sort of comment, what do you think it's like to write an operating system like that. So that process that can went through in three weeks because you were I mean you're part of that process, you contributed a lot to uh to UNIX his early development. So what do you think it takes to do that first step, that first kind of from design to reality on the PDP? Well [/INST] let me correct one thing, I had nothing to do with it, so I did not write it. I have never written operating system code </s><s>[INST] um [/INST] and so I don't know now, an operating system is simply code and this first one wasn't very big but it's something that lets you run processes of some let you execute some kind of code that has been written. It lets you store information for periods of time so that it doesn't go away when you turn the power off or reboot or something like that. Um and there's a kind of a core set of tools that are technically not part of an operating system, but you probably need them in this case can wrote an assembler for the PDP seven that worked needed to text editor so that you could actually create text, He had the file system stuff that he had been working on and then the rest of it was just a way to load things, execute able code from the file system into the memory, give it control and then recover control when it was finished or in some other way </s><s>[INST] quit. What was the code written in the primarily the programming language was an assembly PDP [/INST] seven assembler that can created These things were assembly language until probably the call in 1973 or 74, something like that. </s><s>[INST] Uh forgive me if it's a dumb question, but it feels like a daunting task to write any kind of complex system. An assembly absolutely like it feels like impossible to do any kind of what we think of a software engineering. Assembly is to work in a big picture, sort of, [/INST] I think it's hard, it's been a long time since I wrote assembly language. It is absolutely true that assembly language, if you make a mistake, nobody tells you there are no training wheels whatsoever. Um and so stuff doesn't work now. What? And there's nota bookers, Well, there could be de buggers, but that that's the same problem. Right. How do you actually get something that will help you debug it? So part of it is, is um an ability to see the big picture. Now, these systems were not big in the sense that today's pictures. I see the big picture was, in some sense more manageable. I mean, then realistically there's an enormous variation in the capabilities of programmers and ken Thompson who did that first one is kind of the singularity uh in my experience of programmers, uh with no disrespect to you or even to me, he's into now </s><s>[INST] several [/INST] leagues removed. </s><s>[INST] I know there's levels, it's a, it's a fascinating thing that there are unique stars in particular in the programming space and in a particular time, you know, the time matters to the timing of when that person comes along and the wife does have to leave. Like there's this weird timing that happens that and then all of a sudden something beautiful is created. How does it make you feel that there's a system that was created in in three weeks? Or maybe you can even say on a whim, but not really. But of course, quickly that is now you could think of most of the computers in the world run on a UNIX like system. Right? What, how do you enter? Like, uh, if you kind of assume from the alien perspective, if you're just observing Earth that all of sudden these computers took over the world and they started from this little initial seed of UNIX, how does that make you feel? [/INST] It's quite surprising. And you asked earlier about prediction, The answer is no, there's no way you could predict that kind of evolution. And I don't know whether it was inevitable or just a whole sequence of blind luck, I suspect more of the latter. And um, so I look at it and think gee that's kind of neat. I think the real question is what this can think about that because he's the guy arguably from whom it really came tremendous contributions from Dennis Ritchie and then others around in that Bell Labs environment, but you know, if you had to pick a single person that would be can, </s><s>[INST] so you've written in your book UNIX a history and a memoir, are there some memorable human stories, funny or profound from that time to just kind of stand out? [/INST] Oh, there's a lot of them in a sense. And again, it's a question of, can you resurrect them in real time when his memory fails? But I think part of it was the Bell Labs at the time was was a very special kind of place to work because there were a lot of interesting people and the environment was very, very open and free. It was a very cooperative environment, very friendly environment. And so if you had an interesting problem, you go and talk to somebody and they might help you with the solution. And um and it was a kind of a fun environment to in which people did strange things and um often tweaking the bureaucracy in one way or another, </s>

<s>[INST] what do you think the future programming looks like? What do you hope the future programming looks like if we zoom in on the computational fields on data science machine learning. [/INST] I hope Swift is successful because the goal is swift. The way chris Latner describes it is to be infinitely hackable and that's what I want. I want something where me and the people I do research with and my students can look at and change everything from top to bottom. There's nothing mysterious and magical and inaccessible. Unfortunately with python, it's the opposite of that because python is so slow, it's extremely unhappy. Kable, you get to a point where it's like okay from here on down at sea, so your body doesn't work in the same way your profile, it doesn't work in the same way your build system doesn't work in the same way. It's really not very hackable. </s><s>[INST] What's the part you would like to be hackable? Is it for the objective of optimizing training of neural networks in France. The neural networks is a performance of the system or is there some non performance related? Just? [/INST] It's everything. I mean in the end I want to be productive as a practitioner. So that means that so like at the moment our understanding of deep learning is incredibly primitive. There's very little we understand most things don't work very well even though it works better than anything else out there. There's so many opportunities to make it better. So you look at any domain area, like I don't know, speech recognition with deep learning or natural language processing classification with deep learning or whatever every time I look at an area with deep learning. I always see like, oh it's terrible. There's lots and lots of obviously stupid ways to do things that need to be fixed. So then I want to be able to jump in there and quickly experiment and make them better. </s><s>[INST] Do you think the programming language is has a role in that [/INST] huge role? Yeah. So currently python has a big gap in terms of our ability to, uh, innovate particularly around recurrent neural networks and, um, natural language processing because because it's so slow the actual loop where we actually looked through words, we have to do that whole thing in CUDA C. So we actually can't innovate with the colonel the heart of the most important algorithm. And it's just a huge problem and this happens all over the place. So we hit, you know, research limitations. Another example, convolutional neural networks which are actually the most popular architecture for lots of things. Maybe most things in declining, we almost certainly should be using sparse convolutional neural networks. Um, but only like two people are because to do it, you have to rewrite all of that courtesy level stuff. And yeah, this researchers and practitioners don't so like there's just big gaps in like what people actually research on what people actually implement because of the programming language problem. </s><s>[INST] So, you think, uh, do you think it's it's just too difficult to write in code to see, uh, that programming like a higher level programming language like Swift should enable the the easier fooling around creative stuff with our own ends or was part competition, you know, it's kind of who, who, who's uh who's at fault, who's, who's in charge of making it easy for a researcher to player. [/INST] I mean no one's fault, he's got around to it yet. It's just it's hard. Right? And I mean part part of the fault is that we ignored that whole ap l kind of direction most nearly everybody did for 60 years, 50 years. But recently people have been starting to reinvent pieces of that and kind of create some interesting new directions in the compiler technology. So the place where that's particularly happening right now is something called lia, which is something that chris Lattner, the Swift guy is leading and because it's actually not going to be swift on its own, that solves this problem because the problem is that currently writing, hey, acceptably fast Gpu program is too complicated regardless of what language you use. And that's just because if you have to deal with the fact that I've got, you know, 10 0 threads and I have to synchronize between them all. And I have to put my thing into grid blocks and think about what's and all this stuff. It's just, it's just so much boilerplate to do that. Well you have to be a specialist at that and it's going to be a year's work too, you know, optimized that algorithm in that way. But with things like tensor comprehension, Xyz and tile and I R and T v m there's all these various projects which are all about saying let's let people create like domain specific languages for tensor computations, these are the kinds of things we do are generally on the Gpu for deep learning and then have a compiler which can optimize that tensor computation. A lot of this work is actually sitting on top of a project called Halide, which was is a mind blowing project where they came up with such a domain specific language. In fact to one domain specific language for expressing. This is what my tensor computation is. And another domain specific language for expressing. This is the kind of the way I want you to structure The compilation of that. Like do it block by block and do these bits in parallel and they were able to show how you can compress the amount of code by 10 x compared to optimized Gpu code and get the same performance. So that's like so these other things are kind of sitting on top of that kind of research and mm HLA is pulling a lot of those best practices together. And now we're starting to see work done on making all of that directly accessible through swift so that I could use swift to kind of write those domain specific languages and hopefully we'll get them swift cuda kernels written in a very expressive and concise way that looks a bit like jay and apple and then swift layers on top of that. And then I switched UI on top of that, you know, it'll be so nice if we can get to that point, </s>

<s>[INST] So there you have to call it human [/INST] nature at A level and you need to just be able to ingest many many more sources of data that are on wildly different time scales, right? So if I'm an HFT my high frequency trader like I don't, I don't, I really my main source of data is just the data from the exchanges themselves about the activity in the exchanges, right? And maybe I need to pay, you know, you need to keep an eye on the news, right? Because you know, that can cause sudden, you know, the the, you know, ceo gets caught in a scandal or you know, gets run over by a bus or something that can cause very sudden changes. But you know, I don't need to understand economic cycles. I don't need to understand recessions, I don't need to worry about the political situation or war breaking out in this part of the world because you know, all you need to know is as long as that's not gonna happen In the next 500 milliseconds, then, you know, my model is good when you get to these longer time scales, you really have to worry about that kind of stuff. And people in the machine learning community are starting to think about this. We we held a we we jointly sponsored a workshop at penn with the Federal Reserve Bank of philadelphia a little more than a year ago on, you know, I think the title is something like machine learning for macroeconomic prediction. Um you know, macroeconomic, referring specifically to these longer time scales and you know, it was an interesting conference, but it, you know, my it left me with greater confidence that you have a long way to go to, you know, and so I think that people that, you know, in the grand scheme of things, you know, if, if somebody asked me like, well who's job on Wall Street is safe from the bots. I think people that are at that longer time scale and have that appetite for all the risks involved in long term investing and that really need kind of not just algorithms that can optimize from data, but they need views on stuff, they need views on the political landscape, economic cycles and the like. Um and I think, you know, they're they're they're pretty safe for a while as far as I can </s><s>[INST] tell. So Warren Buffett's [/INST] jobs, Yeah, I'm not seeing, you know, a robo. Warren buffet any time. So </s><s>[INST] give him comfort. Last question, If you could go back to, if there's a day in your life you could relive because it made you truly happy, maybe you're outside family, [/INST] well, otherwise, you know, </s><s>[INST] what, what day would it be, what can you look back? You remember just being uh profoundly transformed in some way or? Yeah, blissful. I'll [/INST] answer a slightly different question which is like, what's this a day in my my life or my career that was kind of a watershed moment. I went straight from undergrad to doctoral studies and you know, that's not at all a typical and I'm also from an academic family, like my dad was a professor, my uncle on his side as professor. Both my grandfathers were professors, all </s><s>[INST] kinds of majors to philosophy and [/INST] yeah, they're kind of all over the map. Yeah. And I was a grad student here just up the river at Harvard and came to study with less valiant, which was a wonderful experience. But you know, I remember my first year of graduate school I was generally pretty unhappy and I was unhappy because you know at Berkeley as an undergraduate, you know, Yeah, I studied a lot of math and computer science but it was a huge school first of all and I took a lot of other courses as we discussed, I started as an english major and took history courses and art history classes and had friends, you know, that did all kinds of different things and you know, Harvard's a much smaller institution than Berkeley and its computer science department especially at that time was was a much smaller place than it is now and I suddenly just felt very, you know like I had gone from this very big world to this highly specialized world um and now all of the classes I was taking were computer science classes and I was only in classes with math and computer science people and so I was, you know, I thought often in that first year of grad school about whether I really wanted to stick with it or not and um you know, I thought like, oh I could, you know, stop with the masters, I could go back to the Bay Area and to California and you know, this was one of the early periods where there was, you know, like you could definitely get a relatively good job paying job at one of the one of the tech companies back, you know, that we're the big tech companies back then. And so I I distinctly remember like kind of a late spring day when I was kind of, you know, sitting in boston common and kind of really just kind of chewing over what I wanted to do with my life. And I realized like, okay, you know, and I think this is where my academic background helped me a great deal, I sort of realized, you know, yeah, you're not having a great time right now, this feels really narrowing, but you know, that you're here for research eventually and to do something original and to try to, you know, carve out a career where you kind of, you know, choose what you want to think about, you know, and have a great deal of independence. And so, you know, at that point, I really didn't have any real research experience yet. I mean, it was trying to think about some problems with very little success, but but I knew that, like, I hadn't really tried to do the thing that I knew I'd come to do, and so I thought, you know, I'm gonna, I'm gonna stick to, you know, stick through it for the summer and you know, and and and that was very formative because I went from kind of contemplating quitting too, you know, a year later, being very clear to me, I was going to finish because I still had a ways to go, but I kind of started doing research. It was going well, it was really interesting and it was sort of a complete transformation. You know, it's just that transition that I think every doctoral student makes at some point, which is just sort of go from being like a student of what's been done before to doing, you know, your own thing and figure out what makes you interested in what your strengths and weaknesses are as a researcher. And once, you know, I kind of made that decision on that particular day at that particular moment in boston common. Um you know, I'm glad I made that decision </s>

<s>[INST] and there wasn't really a plan because I got a chance to talk to the ceo of Qualcomm recently and he, his advice is always have a plan and it sounds [/INST] like you're </s><s>[INST] saying you [/INST] don't have a </s><s>[INST] plan, don't need to have a plan. Just listen to your gut, your passion and follow that and see see where that takes you because it's telling you something. [/INST] Yeah, I think you know, I guess the plan could be specific or could be as general as I just want to go far away and see something very different. That's my plan and I did </s><s>[INST] just just [/INST] followed my nose um from one thing to the next. Just being interested, following my passion and again, very fortunate I could do that. </s><s>[INST] Are there places in the world you're kind of thinking about [/INST] that </s><s>[INST] your life might take you at some point just um to be a doctor there for a time to explore for a time. They haven't yet. [/INST] You know, I have some colleagues who do kind of global health work in various countries in africa and central and South America. I would really love to um go to some of those places, not just for a short trip but hopefully for an extended period of time um with sort of the health care being the ticket in. Um but then maybe even bringing my Children or just you know, I guess at this point some of the travel I dream about is sort of replicating what I did and showing it to my kids in a way. Um but there's still a lot I haven't seen and would love to see as well, but um I think those opportunities sort of lend themselves well. Um you know, as a doctor with kind of the ability to go there and sort of help help patients, but also teach medical students and residents. Teaching is actually a huge part of being a doctor that's underappreciated, but that's actually part of the fun of being a doctor is that you're also a teacher. Of course the word doctor means teacher, but it's come to mean something else. Um but you know, i in some of my jobs, I'm working alongside medical students and residents and I'm giving them my knowledge, my wisdom sharing with them stories. And so that's a very satisfying part of the job. </s><s>[INST] If we can take a brief step into a dark place together for a time, Is there what is a dark place you've gone in your mind in your life? What would be the darkest place you've ever gone [/INST] for </s><s>[INST] time for a moment. Um and how did you survive? How did you overcome it? [/INST] That's a very good question, I would say. Um I haven't had as dark moments, as many of the people who I care for in the emergency room. I'm fortunate in that way. I've had a pretty, you know, enjoyable, satisfying life, you know, I think everybody um has dark moments though, including me, one of the most shocking things, I feel like becoming an adult. My two big realizations have been one, no one knows what they're doing and to suicide is incredibly common, like in all humans, in all societies and that I just find shocking. I mean I've never seriously contemplated myself, but I wouldn't say it hasn't crossed my mind during some more stressful times of life. Um I think it crosses everyone's mind and it's sort of as a kid, I found that I never would have guessed how common suicide is </s><s>[INST] an important question, sort of the the camus question, like why live, why, why is like life, especially when you're struggling, especially when life is shit, like why am I doing any of this? And then on top of that chemistry of your brain, [/INST] it </s><s>[INST] could be as simple as diet and nutrition and aforementioned exercise and things like this that affect the chemistry such that you're more predisposed to go to the places of asking the question why and maybe struggling to find a good answer because it's actually a question with no good answer except something in your chemistry says, well that kind of like it, but there's no good intellectual answer and especially if day to day it's pain. You get to see these stories of um you know, Robin Williams, these people that are on top of the world from an external perspective, but from an internal perspective, it's struggle every day is pain feels hopeless. And um yeah, that's a that's a question we all have to struggle with or learn how to ignore. Maybe because if you ask the question too much, you're not going to, you're not going to find a good answer. That's a choice you make. I personally think you should ask that question a lot. Um [/INST] but </s><s>[INST] maybe because I have the luxury of the chemistry I have where I'm not in danger of seriously contemplating suicide, but why lives an important question to answer constantly and struggle to answer that constantly. But people I am uh I've gotten I've been extremely fortunate to meet people over the past couple of years that are really struggling. And you have probably met mm people who are really struggling like orders of magnitude more people who are really struggling. Some of it is psychological. A lot of it is biological and man life. Life is a motherfucker. It's pretty tough, [/INST] very true. I do think also past trauma plays a big role there. Like we talked about, you know, war wounds and PTSD and a lot of people grew up, I mean, which is horrific childhoods, they're abused in one way or another, and I think a lot </s><s>[INST] of people who have [/INST] not, not, I'm not saying a majority, but a lot of people for instance who I see in the er coming in for threatening suicide or actually trying and failing and being brought to the er you know, a lot of them just have really traumatic experiences um you know, saw their parent commit suicide uh were abused, you know, these leave scars in the, in the human brain and mind and a lot of the their subsequent lives of whether it's substance abuse, alcoholism, et cetera, is almost trying to escape from their own memories and it's sort of such, this overwhelming battle sometimes, like sometimes people get ruined, it seems, and just can't be fixed, you know what I mean? Yes, you can improve diet and health and your life choices and seek out your passion and exercise and those definitely will help, but sometimes just like, you know, you bear the scars of the past and there's no getting rid of them. </s>

<s>[INST] and it's a little, probably a little bit more expensive to do it the way nature does because nature is uh less sensitive to the loss of the individual and uh cost wise in robotics. I think you're more sensitive to losing individuals. [/INST] I think that's true, although if you look at the price to performance ratio of robotic components, it's it's coming down dramatically, right? It continues to come down. So I think we're as symptomatically approaching the point where we would get, yeah, the cost of individuals will really become insignificant. </s><s>[INST] So let's step back at a high low of you. The impossible question of what kind of as an overview what kind of autonomous flying vehicles are there in general. [/INST] I think the ones that receive a lot of notoriety are obviously the military vehicles, military vehicles are controlled by a base station but have a lot of human supervision. But I have limited autonomy, which is the ability to go from point A to point B. And even the more sophisticated now sophisticated vehicles can do autonomous takeoff and landing </s><s>[INST] and those usually have wings and they're heavy, usually [/INST] their wings, but there's nothing preventing us from doing this for helicopters as well. There, I mean there are many military organizations that have autonomous helicopters in the same vein and by the way, you look at auto pilots and airplanes and it's it's actually very similar. In fact, I can, one interesting question we can ask is if you look at all the air safety violations, all the crashes that occurred would have happened if the plane were truly autonomous. And I think you'll find that many of the cases because of pilot error. We make silly decisions and so in some sense, even in air traffic, commercial air traffic, there's a lot of applications, although we only see autonomy being enabled at very high altitudes when, when the pilot, the plane is on autopilot, </s><s>[INST] there's still a role for the human and that kind of autonomy is uh, you're kind of implying, I don't know what the right word is, but it's a little dump dumber than it could be. [/INST] Right. So, so in the lab, of course we can we can we can afford to be a lot more aggressive. And the question we try to ask is can we make robots that will be able to make decisions without any kind of external infrastructure? So what does that mean? So the most common piece of infrastructure that airplanes use today is GPS GPS is also the most brutal form of information. Um, if you have driven in a city tried to use GPS navigation, tall buildings, you immediately lose GPS um, and so that's not a very sophisticated way of building autonomy. I think the second piece of infrastructure that rely on his communications. Again, it's very easy to jam communications. In fact, if you use wifi, you know that wifi signals dropout cell signals drop out. So to rely on something like that is not is not good. The third form of infrastructure we use and I hate to call it infrastructure, but it is that in the sense of robots as people. So you can rely on somebody to pilot you. Uh And so the question you want to ask is, if there are no pilots, there's no communications with any base station, if there's no knowledge of position and if there's no a priori map opera, only knowledge of what the environment looks like operate a model of what might happen in the future can robots navigate. So that is true autonomy. So </s><s>[INST] that's that's true autonomous. And we're talking about, you mentioned like military applications of drones. Okay, so what else is there? You talk about agile autonomous flying robots, aerial robots. So that's a different kind of, it's not winged, it's not big, at least it's small. [/INST] So I used the word agility mostly. Or at least we're motivated to do agile robots mostly because robots can operate and should be operating in constrained environments. And if you want to operate the way a global hawk operates, I mean the kinds of conditions in which you operate are very, very restrictive. If you go, I want to go inside a building, for example, for search and rescue or to locate an active shooter or you want to navigate under the canopy in an orchard to look at health of plans or to look for the count, to count fruits, to measure the tree, the tree trunks. Um, These are things we do by the way, </s><s>[INST] uh, some cool agriculture stuff you've shown in the past. It's really all right. [/INST] So, so in those kinds of settings you, you do need that agility, agility does not necessarily mean you break records for 100 m dash what it really means is you see the unexpected and you're able to maneuver uh in a safe way and in a way that that gets you the most information about the thing you're trying to do </s><s>[INST] by the way. You may be the only person who in a ted talk has used a math equation, which is amazing. People should go see one of your actually, [/INST] it's very interesting because the Ted curator chris Anderson told me you can't show math and you know, I thought about it, but that's who I am. I mean that's that's what that's our work. And so I felt compelled to give the audience a taste for at least some math. </s><s>[INST] So on that point, simply, what does it take to make a thing with four motors fly a quadcopter, one of these little flying robots? How hard is it to make it fly? How do you coordinate the four motors? What's how do you convert this? Those motors into actual movement? [/INST] So, this is an interesting question. We've been trying to do this since 2000. It is a commentary on the senses that were available back then. The computers that were available back then and a number of things happened Between 2000 and 2007. one is the advances in computing, which is, and so we all know about moore's law. But I think Uh, was a tipping point in the year of the iPhone, the year of the cloud. Uh, lots of things happened in 2007. Um, but going back even further inertial measurement units as a sensor really matured. Again, lots of reasons for that. Um, certainly there's a lot of federal funding, particularly DARPA in the US, but they didn't anticipate this boom uh, in I am us. Um, but if you look uh subsequently what happened is that every, every car manufacturer had to put an airbag in, which meant you had to have an accelerometer on board. So that drove down the price to performance ratio, </s>

<s>[INST] said that the world's first trillionaire will be somebody in climate change in the past. Um, let's update that. What's today as we stand here today, What sector will the world's first trillionaire come from? [/INST] I think its energy transition, So </s><s>[INST] energy. So the things we've been talking [/INST] about really, </s><s>[INST] so isn't it? Okay, well, [/INST] I think, I think the way that I think about, </s><s>[INST] so this is a single individual, Sorry to interrupt you see their ability to actually build a company that makes a huge amount of money as opposed to this distributed idea that you've been talking to. [/INST] I'll give you my philosophy on </s><s>[INST] wealth. [/INST] Um, most of it is not you, an enormous amount of it is the genetic distribution of being born in the right place and blah blah blah irrespective of the boundary conditions of how you were born or where you were raised, right? So, you know, at the end of the day, you and I ended up in the United States, it's a huge benefit to us. 2nd is the benefit of our age. It's much better and much more likely to be successful as a 46 year old in 2023 than a 26 year old in 2023. Because in my case, I have demographics working for me For the 26 year old. He or she has demographics working slightly against them. Can </s><s>[INST] you explain that a little bit? What are the demographics here [/INST] in the case of me, the distribution of population in America looks like a pyramid and in that pyramid, I'm sweaty edged in between these two massive population cohorts, the boomers and then these, you know, gen Z and millennials. Um, and that's a very advantageous position. It's not dissimilar to the position that Buffett was, where he was, you know, packs packaged in between boomers beneath him and the silent generation above him and being in between two massive population cohorts turns out to be extremely advantageous because when the cohort above you transitions power and capital and all of this stuff, you're the next person that likely gets handed it. So we have a disproportionate likelihood to be, you know, we are lucky to be older than younger. Um, so that's, that's an advantage. And then the other advantage that has nothing to do with me is that I stumbled into technology. I got a degree in electrical engineering and I ended up coming to Silicon Valley. And it turned out that in that moment it was such a transformational wind of change that was at my back. Right? So the wealth that one creates is a huge part of those variables. And then the last variable is your direct contributions in that moment. And the reason why that can create extreme wealth is because when those things come together at the right moment, it's like a chemical reaction, I mean it's just crazy. So that was sort of part number one of what I wanted to say. The second thing is when you look then inside of these systems where you have all these tailwinds, right? So in tech, I think I benefit from these three big tail winds. If you build a company or a part of a company or a part of a movement, your economic participation tends to be a direct byproduct of the actual value that that thing creates in the world and that the thing that that creates in the world will be bigger if it is not just an economic system, but it's like a philosophical system, it changes the way the governance happens, it changes the way that people think about all kinds of other things about their lives. So there's a reason, I think why database companies are worth X. Social companies are worth y but the military industrial complex is worth, you know, as much. And I think there is a reason why that if you for example, were to go off and build some newfangled source of energy that's clean and hyper abundant and safe. That what you're really going to displace or reshape is trillions and trillions of dollars of worldwide GDP. So the global GDP is I call it 85 trillion, right? It's going at 2-3% a year. So in the next 10 years we'll be dealing with $100 trillion of GDP, right? Somebody who develops clean energy in 2035 will probably shift 10% of that around $10 trillion. A company can easily capture 30% of the market, $3 trillion. A human being can typically own a third of one of these companies. $1 trillion. So you can kind of get to this answer where it's like it's going to happen in our lifetime. But you have to, I think find these systems that are so gargantuan and they exist today. It's more bounded because price discovery takes longer and an existing thing, it's more unbounded because you know what it is. You know, the tentacles that energy reaches Right of that $80 trillion 60 of the 80. </s><s>[INST] Do you have an idea of which energy, uh, which alternate energy, sustainable energy is the most promising. Well, [/INST] I think that we have to do a better job of exploring what I call the suburbs of the periodic table. So, you know, we're really good in Seattle, you know, the upper northwest, you know, we're kind of good in Portland but we're not existent in san Diego. And we have zero plan for north Carolina through florida. Yeah. And so </s><s>[INST] is that a fancy way of saying nuclear is should be part of the discussion. [/INST] I think nuclear, I mean room temperature semiconductors, I'm not convinced right now that the existing set of nuclear solutions will do a good job of scaling beyond bench scale. I think there is a lot of complicated technical problems that make it work at a bench scale level even partially, but the energy equation is going to be very difficult to overcome in the absence of some leaps in material science. </s>

<s>[INST] Is it possible to sort of optimally do this kind of selection [/INST] to </s><s>[INST] do this election about um of the equilibrium year chasing? So is there like a meta problem to be solved here? [/INST] So the meta problem is, in some sense, um how do you, how do you understand the nash equilibrium that the other players are going to play? Um and and even if you do that again, there's no guarantee that you're going to win. So, you know, if you're playing, uh if you're playing risk, like I said, and and all the other players decide to team up against you, you're gonna lose nash equilibrium doesn't help you there. And so there was this big debate about whether nash equilibrium and all these techniques that compute it are even useful once you go outside of two player zero sum games. Now, I think for many games, there is a valid criticism here. And I think when we talk about when we go to something like diplomacy, we run into this issue that the approach of trying to approximate a nash equilibrium doesn't really work anymore, but it turns out that in six player poker, um because six player poker is such an adversarial game um where none of the players really try to work with each other. The techniques that were used in two player poker to try to approximate an equilibrium. Those still end up working in practice in in six player poker as </s><s>[INST] well. There's some deep way in which six player poker is just a bunch of heads up poker, like games in one, it's like, it's like embedded in it. So the competitiveness um, is more fundamental to poker than the cooperation, [/INST] right? Yeah, poker is just such an adversarial game. There's no real cooperation. In fact, you're not even allowed to cooperate in poker, it's considered collusion, it's against the rules. Um and so for that reason, the techniques end up working really well and I think that's true more more broadly and extremely adversarial games in general, </s><s>[INST] but that's sort of in practice versus being able to prove something. [/INST] That's right. Nobody has a proof that that's the case. And it could be that, that six player poker belongs to some class of games where approximating a nash equilibrium through self play provably works well. Um and, you know, there are other classes of games beyond just two player zero somewhere. This is proven to work well. So there are these kinds of games called potential games, which I won't go into, It's kind of like a complicated concept, but um there are classes of games where uh, this approach to approximating a nash equilibrium is proven to work well now six player poker is not known to belong to one of those classes. But it is possible that there is some classic games where it either provably performs well or provably performs not that badly. </s><s>[INST] So what are some interesting things about pluribus that was able to achieve human level performance on this or super human level performance on the six player version of poker [/INST] personally. I think the most interesting interesting thing about this is that it was so much cheaper than laboratories. I mean laboratories, if you had to put a price tag on on the computational resources that went into it, I would say the final training run took about $100,000. You go to pluribus. The final training run would cost like less than $150 on aws is </s><s>[INST] just normalized to computational inflation. So meaning uh this is this just does this just have to do with the fact that was trained like a year later? [/INST] No, no, no it's not. It's I mean first of all like yeah computing resources are are getting cheaper every day and like but you're not going to see 1000 fold decrease in the computational resources over two years. Um or even anywhere close to that. The real improvement was algorithmic improvements and in particular the ability to do deep lamented search. So </s><s>[INST] it does depth limited search also work for LA. [/INST] Yeah. Yes. So where this dedicated search came from is you know, I developed this technique and um ran it on two player poker first and that reduced the computational resources needed to make an ai that was superhuman from, you know, $100,000 for the broadest to something you could train on your laptop. What </s><s>[INST] do you learn from that? Um from that discovery, [/INST] what I would take away from that is that algorithmic improvements really do matter </s><s>[INST] how would you describe the more general case of limited depth search? So it's basically constraining the scale temporal or in some other way of the computation you're doing in some clever way. So like with like how else can you significantly constrain computation? Right. Well, [/INST] I think the idea is that we want to be able to leverage search as much as possible and the way that we were doing it in the broad is required us to search all the way to the end of the game. Now, if you're playing a game like chess, the idea that you're gonna search always to the end of the game is kind of unimaginable, right? Like there's just so many situations where you just won't be able to use search in that case or the cost would be um you know, prohibitive and this technique allowed us to leverage search and without having to pay such a huge computational cost for it and be able to apply it more broadly. </s><s>[INST] So to what degree did you use neural nets for laboratories and pluribus and more generally, what role do neural nets have to play in? Um in super human level performance in poker? [/INST] So we actually did not use neural nets at all for libraries or play this. And a lot of people found this surprising back in 2017, I think they found it surprising today um that we were able to do this without using any neural nets. Um and I think the reason for that, I mean I think neural nets are um incredibly powerful and the techniques that are used today, even for poker eyes do rely quite heavily on neural nets. Um but it wasn't the main challenge for poker, like I think what neural nets are really good for. If you're in a situation where finding features for a value function is really difficult, then neural nets are really powerful. And this was the problem in go right. Like the problem in go was that for the final problem and go at least was that nobody had a good way of looking at a board and figuring out who was winning or describing um through a simple algorithm who was winning or losing. And so their neural nets were super helpful because you could just feed in a ton of different board positions into this neural net and it would be able to predict then who was winning or losing. But in poker, the features weren't the challenge, the challenge was. How do you design a scalable algorithm that would allow you to find this balance strategy? Um that would understand that you have to bluff with the right probability. </s>

<s>[INST] and then so these passions and underneath it is the is the belief that you can be anything you want [/INST] to, you can you know, I told my kids this, you know, when they were young, you know, it was tough especially for my son. So when it was about 56 years we knew Nate was color blind. You know my my wife's brothers are both color blind. It's really colored deprived, colored, applying to see black and white. He can't tell he has issues with greens, reds, browns. It's funny if you're ever around someone like that because I'll go, I'll go, what are you looking at? He goes right over there by the red thing. I'm like, what are you looking at? Go this like he had a hat on one day, I know which one you're getting. He had a hat in his hand, it was green. He goes, I'm gonna get the green when I go. Oh this one right here. He goes, no, the one on my head, I go Nate, that one's brown. He's like, leave me alone dad. He got the brown hat because to him it looked great. So he couldn't fly. He came, he said, I know what you want to do not, you know, you're talking to your kids and you know what you want to use and I want to be a pilot? No, no, I got to tell him because he's looking at me because I'm a pilot. Do you can't be a pilot, he's like why can't I be a pilot? I said because you you got issues, You know so you got to redirect and the other one was because I had I stopped flying, I was 42 years old and I was like and it was my childhood dream so it's like a pro athlete. I know exactly what it feels like when you know Brett Farve has to walk away from the NFL when you still can do it, </s><s>[INST] good choice of quarterback by the way the greatest of all time but [/INST] whatever so you do when you look at it and you go I understand what those guys feel like when you have to walk away from something you love and you think you can still do it. Um so I told them, I said look I was talking to both my kids and I said you know find something that you want to do that you love to do and that you can do your whole life and you should be able to do good things for other people, you want to be able to help other people. That's what I said. So both of my kids and there's no one in my family, both of my Children, one of them is my daughter is a doctor doing a residency in internal medicine right now and my son is in his third year and they're both gonna be doctors and until I look at it is you know people go oh you got two dogs and I don't care. I told my kids if you want to be a garbage man or you want to dig ditches, I don't care, just be the best ditch digger that you can be. I said I'd be happy doing it because what you also find is that we are in this big pursuit of money, money, money, money, money, money, money, that's what makes the world go round. But what you realize and I'll go back to my grandfather who didn't have a lot of money And he was probably one of the most happy people on life and unfortunately he died and he died at 65 had a massive heart attack because he didn't tell that, he he kind of knew it was happening and he just made the choice to do and it was devastating to the entire family, but he didn't, he didn't have a lot of money, but I'll tell you what, I know a lot of rich people who have funerals and there's nobody at them. And my grandfather who was a beer delivery guy had I it literally it was like three miles long. It was crazy. Yeah, who died. That was because it was like he's a Catholic, you know Italian, he goes, You know who died, the pope and I'm going to ask my grandfather and then the next day and I went to was my aunt, his sister and it was like, you know 30 people and I looked at my mother and I said, where's everybody at? She goes, oh no, this is normal. This is what a normal funeral looks like. So it's, you know, for young kids, bottom line one, be nice kindness will get you. I'm a big believer in karma, kindness will get you a long way in the world. You know, it's easy, It's easy to be nice. It doesn't cost you anything. I said, you know, and get rid of the hate. And number two is follow your dreams because everyone is capable of everything. And there's a, there's a self realism, like, you know, if you really have trouble with math, getting a PhD in applied math. It's probably not something you're going to be able to do, but understand yourself what your own capabilities are and you know, inside your heart, don't let anyone ever tell you what you can and can't do. You have to determine that yourself and go for it and and you can do anything. It's just, it's it's a great the world's incredible. It really is. </s>

<s>[INST] So, but the testers to vectorized sort of formal representation, that's meaningful thing. And your intuition now, having worked with all this media is that once you are able to form that representation, you can basically take anything's any sequence. Uh Is there going back to Starcraft is their limits on the length? Uh So we didn't really touch on the long term aspect. How did you overcome the whole really long term aspect of things here? Is there some tricks or So, [/INST] the main streak. So Starcraft, If you look at absolutely every frame, you might think it's quite a long game. So we would have to multiply 22 times 60 seconds per minute times, maybe at least 10 minutes per game on average. So there are quite a few frames but the trick really was to only observe, in fact which might be seen as a limitation, but it is also a computational advantage only observe when you act and then what the neural network decides is what is the gap going to be until the next action. Um and if you look at most Starcraft games that we have in the data set that Blizzard provided, it turns out that most games are actually only, I mean it is still a long sequence, but it's maybe like 1000 to 1500 actions, which if you start looking at LST M'S large, LSD Ems transformers, it's not like it's not that difficult, especially if you have supervised learning. If you had to do it with reinforcement learning the credit assignment problem, what is it that in this game that made you win? That would be really difficult. But thankfully because of imitation learning, we didn't have to deal with this directly, although if we had to, we tried it and what happened is you just take all your workers and attack with them and that sort of is kind of obvious in retrospect because you start trying random actions. One of the actions will be a worker that goes to the enemy base and because itself play, it's not going to know how to defend because it basically doesn't know almost anything and eventually what you develop is this take our workers an attack because the credit assignment issue in a rally is really, really hard. I do believe we could do better and that's maybe research challenge for the future. Um but yeah even even in starcraft the sequences are maybe 1000 which I believe there is within the realm of what transformers can do. </s><s>[INST] Yeah I guess the difference between Starcraft and go is uh in go and chest stuff starts happening right away. So there's not yeah it's pretty easy to self play. Not easy but to self play as possible to develop reasonable strategies quickly as opposed to Starcraft. [/INST] Mean in go there's only 400 actions but one action is what people would call the God action. That would be </s><s>[INST] if you had [/INST] expand the whole third street. That's the best action. If you did mini max or whatever algorithm you would do if you had the computational capacity. But in Starcraft the 400 is minuscule. Like in 400 you don't even like you couldn't even click on the pixels around a unit. Right. So um I think the problem there is in terms of action, space size is way harder. So and that search is impossible. So there's quite a few challenges indeed that make this um kind of a step step up in terms of machine learning for humans. Maybe they're playing Starcraft it seems more intuitive because it looks real. I mean you know like the graphics and everything moves smoothly. Whereas I I don't know how to I mean go is a game that I wouldn't really need to study. It feels quite complicated, but for machines kind of maybe </s><s>[INST] it's the reverse. Yes, which shows you the gap actually between deep learning and however the heck brains work. Uh So you developed a lot of really interesting ideas. It's interesting to just ask what's the, what's your process of developing new ideas? Do you like brainstorming with others? Do you like thinking alone? Do you like uh uh like was it Ian Goodfellow said he came up with Ganz after a few beers. Uh he thinks beers are essential for coming up with new ideas. [/INST] We had beers to decide to play another game game of Starcraft after a week. So it's really similar to that story actually, I explain this, you know, in a deepmind retreat and I said this is the same as the gun story. I mean we were in a bar and we decided let's play again next week and that's what happened. </s><s>[INST] I feel like we're getting the wrong message to young undergrads. Yeah, but in general like yeah, like brainstorming, do you like thinking alone, working stuff out? And so [/INST] I think, I think throughout the years also things changed, right? So initially I was very fortunate to be um with great minds like Geoff Hinton, Jeff Dean aliases keeper. I was really fortunate to join Brain at a very good time. So at that point, ideas, I was just kind of brainstorming with my colleagues and learned a lot and keep learning is actually something you should never stop doing, right. So learning implies reading papers and also discussing ideas with others. It's very hard at some point to not communicate that being reading a paper from from someone or actually discussing, right? So definitely that communication aspect needs to be there whether it's written or oral. Um nowadays I'm also trying to be a bit more strategic about what research to do. So I was describing a little bit this sort of tension between research for the sake of research and then you have on the other hand, applications that can drive the research, right? And honestly the formula that has worked best for me is just find a hard problem and then try to see how research fits into it, how it doesn't fit into it and then you must innovate. So I think machine translation drove sequence to sequence. Um then maybe like learning algorithms that had to like communitarian algorithms led to point our networks, um Starcraft led to really scaling up imitation learning and the alpha star league. So that's been a formula that I personally like, but the other one is also about it and I seen it succeed a lot of the times where you just want to investigate model based RL as as kind of a research topic and then you must then start to think well how are the tests, how are you going to test these ideas. Um you need to kind of a minimal environment to try things. You need to read a lot of papers and so on and that's also very fun to do and something I've also done quite a few times, both brain and the mind and obviously as a PhD. So, so I think besides that the ideas and discussions, I think it's important also because you start sort of guiding not only your own goals, but other people's goals um, to the next breakthrough. So you you must really kind of understand these, you know, feasibility. Also, as we were discussing before, right, whether whether this domain is ready to be tackled or not and you don't want to be too early, you obviously don't wanna be too late. So it's it's really interesting um this strategic component of research, which I think as a grad student, I just had no idea. You know, I just read papers and discussed ideas and I think this has been maybe the major change and I recommend people kind of fit forward to success how it looks like and try to backtrack other than just kind of looking out. This looks cool, this looks cool. And then you do a bit of random work which sometimes you stumble upon some interesting things, but in general, it's also good to plan a bit. </s>

<s>[INST] embodies metaphor. [/INST] Yes, he embodies metaphor. He's the gathering. We call him the gathering metaphor bobby. The gatherer metaphor </s><s>[INST] is a metaphor for something else for somebody else's life. Someday he'll be in the dictionary. [/INST] Yeah, representing </s><s>[INST] some kind of concept. Maybe the metaphor itself [/INST] once, once this whole thing falls apart and we are climbing the kudzu vines that spiral up the Sears tower, like they say in fight club bobby will go back to his gathering form and be happy as a pig and shit. Just walking around the loincloth with his bird hanging out, cracking jokes to people and climbing up on them for a stool, lap dancer, Whatever. I'd </s><s>[INST] love to dig into something. He he did. You guys did a lot of great podcast together. He asked you in a very uncomfortable process of why you don't do stand up. So let me ask you, do you hate money. [/INST] Well I'm originally from Canada. Yeah, so I'm uh yeah, I'm I'm a freaking pinko socialist. Is that what where you come from? That's not a nice thing to say, </s><s>[INST] I thought the soviet union, that is a nice thing to say, like [/INST] a </s><s>[INST] is a good he's a good socialist. [/INST] Yeah, </s><s>[INST] red, [/INST] like </s><s>[INST] bold colors. Yeah, there was an interesting tension in your voice and the way you talked about it, there's just not a source of happiness for you. You you you respect the art form, but it was not something that you were connected to you. He felt connected to. [/INST] That's a good way to put it. Yeah, I, I respect the art form a lot. I, and I grew up with all the albums and stuff. I had an older brother and sister who so I, you know, we had the we had George Carlin, we had, you know, Richard Pryor, we had robert klein, we had Gilda live the Gilda Radner concert, we had, we had all sorts of stuff, but you know, I don't know, there's a lot of, there's a lot of reasons, I do feel like a career in show business Is, you know, they, it never goes the way you plan, like most things and I was fortunate enough to get started outside of my native Vancouver or in my native Vancouver, I grew up in the burbs outside and there was a lot of industry there, so I was fortunate enough to get started as an actor when I was like 16. So there there yeah, there was, there were some times early on where I came up with some stand up stuff and did it, but yeah, I quickly abandoned it and then, you know, you go through, you do mad tv and stuff and then and that's where my this is gonna sound weird. Do I sound as anxiety as I did when I was on bobby's podcast trying to avoid his questions? </s><s>[INST] Well he was giving you this face this whole time that was making the whole just atmosphere feel full of anxiety. So I'm trying not to give you the face the whole time I'm saying play cool, play cool, [/INST] okay uh Cool. You said it out loud a couple of times. I did, you know, you cut that out, play cool, play cool. Cut </s><s>[INST] out, cut it out, [/INST] maintain bro, here's what I'll say, there's two ways to do it. I think it's lame when someone who's done one thing for a while goes and starts doing stand up out of nowhere because I think it's an art form that's uh under attack because it's not like anything else you need. Although now you can of course, you know, make whatever you want, it's the era of self publishing, as far as making a product and putting it out there, which is getting easier of course, and I can't wait to talk to you about that with with ai and how it's changing art. Um but the in stand up, all you need is a is a microphone and you know perhaps it would be good to have some mental illness and then you can just run up there and uh, talk forever. And I say this to, to, you know, comedians, it's like, you guys have to deal with Justin influx of people who aren't sure why they're doing comedy. I would ask comedians and like, I mean, not good ones and good ones, you know what they're doing? But everyone else like, what are you doing? Why why you doing stand up? Having said that? I am allergic to money. </s><s>[INST] Yeah. Do you think they have a good answer for that? Why are they doing it? Because I actually, like when I'm in Austin, I like going to open mikes, just listening. It's inspiring to me both the funny and the unfunny people because they've been doing it for several years, sometimes over a decade and they're still at it, they're still right there, there's going for the punch and then especially open mics that are really sad and that there's, you know, only like five other people in the audience and they're usually just other comedians and they're still going all out as if they're in front of a stadium. But [/INST] that, to me sounds like someone who loves it. Yeah, I got no questions for that person. I got questions for someone who goes sideways from here, I'm recognizable doing something and then I'm doing stand up because it's like, and truly, look, I, I've been, I've been fortunate enough to be in the business for a long time, and at this point, if I came up, I mean, doing live stuff is fun. I have friends that are like, um, you know, some guys who are primarily sketch people or you would look at them and sketch people and they can sell tickets for being sketch people and they, and we'll talk about it and they're like, well, you know, I do a monologue and I do a little stand up, then I do a song, then I do another monologue and I play off the audience, do a little stand up. Um, but stand up is, it's almost like playing music in that, you know, people are going up there playing music, but what band have you been listening to? That's what you're gonna sound like. So it's really, I mean, of course I'm speaking from zero experience, but I've heard it takes years of course to find your own voice stand ups that when they first go up there, they're doing some sort of impersonation of so and so, and so, and so, and you gotta you gotta pop this audience that that's paying and you're gonna get run over by the next person who's coming up and uh it's hard to follow the last person who went up before you. And I, I mean, that is a really hard way to, it's a very, quite a gauntlet to begin to find your voice comedically, </s>

<s>[INST] the following is a conversation with Andrew bustamante former CIA covert intelligence officer and U. S. Air Force combat veteran including the job of operational targeting, encrypted communications and launch operations for 200 nuclear intercontinental ballistic missiles. Andrews over seven years as a C. I. A spy have given him a skill set and a perspective on the world that is fascinating to [/INST] explore. </s><s>[INST] And now a quick few second mention of each sponsor. Check them out in the description is the best way to support this podcast. We got Wealthfront for savings [/INST] element </s><s>[INST] for hydration, better help for mental health Express. VPN for privacy and master class for intellectual inspiration, choose wisely my [/INST] friends and </s><s>[INST] now onto the full ad reads never any ads in the middle. I hate those and the ads I do here up front. [/INST] I do try </s><s>[INST] to make interesting but if you must skip them to your friends please still check out the sponsors. I enjoy their stuff. Maybe you will too. This show is brought to you by a new sponsor called Wealthfront. They do savings and automated investing accounts to help you build wealth and save for the future. It's a beautifully designed and streamlined interface. It's honestly really surprising to me how many financial institutions [/INST] of all kinds </s><s>[INST] on the internet don't have a good interface. It's clunky. I don't understand it. They're taking your money. Obviously it should be frictionless to move your money around anyway. I think I have a lot of trouble with companies that don't do a good job with the interface and Wealthfront does a good job with that. I think it's very important to save for the future [/INST] from </s><s>[INST] a stoic perspective, you should consider that you can die any day now, but from a financial perspective you should consider that if you're lucky, you're gonna live for quite a long time and in that case you want to invest into your [/INST] future, </s><s>[INST] nearly half a million people use Wealthfront. You can go check them out at Wealthfront dot com slash lex and you'll get a $50 bonus with a $500 deposit. This episode is also brought to you by element electrolyte drink mix spelled L M N. T. I really love element, it's one of the things that really brings joy to my day. I drink, I don't know how many, Maybe 5, 6 element packets a day. [/INST] So </s><s>[INST] for low carb diets, all that kind of stuff, it's essential to get sodium, potassium, magnesium and they have a really nice balance of that. I think watermelon salt is the one I go to to just feel refreshed to make my water go down in a way where I really enjoy it. It's the thing that makes all the fast thing I do and if I do carnivore diet or keto diet, it's the thing that makes that really enjoyable and easy. [/INST] Never </s><s>[INST] any headaches, never knew that [/INST] kind of stuff. It's </s><s>[INST] the thing I drink before and this thing I drink after to feel refreshed and to make sure I have enough electrolytes in my system to handle the 10 12 13 15 mile runs that I do in the insane heat of Austin. Get a sample pack for free with any purchase when you go to drink element dot com slash [/INST] flex. </s><s>[INST] This episode is also brought to you by better help spelled H E L P help. They figure out what you need a match with a licensed professional therapist in under 48 [/INST] hours. </s><s>[INST] I recently had an offline conversation with Duncan Trussell. I got a chance to meet him. I'm such a huge fan of Duncan, he's such a beautiful person, he's such a sweetheart. He's both hilarious and kind [/INST] but also </s><s>[INST] is willing to go to the wildest ideas and not just for a little bit [/INST] but a full </s><s>[INST] journey through the wildest possible ideas. It's like I feel like Alice in wonderland every time I [/INST] talk to him </s><s>[INST] anyway we've talked about his time with depression and and the value of therapy. [/INST] It </s><s>[INST] really is something [/INST] especially </s><s>[INST] in your darker moments in life that can really, really help. And it could also help to help avoid those darker [/INST] moments. So </s><s>[INST] it's definitely something you should consider check out better help at better help dot com slash lex and save on your first [/INST] month. </s><s>[INST] This show is also brought to you by express VPN. I've used them for many years primarily because they have a big sexy button that I can press and it can do. The thing is supposed to do and that kind of thing is like the easy button. That kind of thing brings joy to my O C. D [/INST] heart that </s><s>[INST] there are software services out there that do a thing and do it well. But I also use it to protect my privacy on the internet, Make sure that SPS can't collect all of my data even when I go to shady websites, which of course I never, ever [/INST] do </s><s>[INST] in incognito mode in chrome. [/INST] The </s><s>[INST] thing also works super [/INST] fast, which is really, </s><s>[INST] really important on any operating system including Linux. It also works on android [/INST] iphone anything </s><s>[INST] you could imagine, any operating system that I at least know of. It works. It's super [/INST] fast. You can </s><s>[INST] use it to watch shows that are geographically constrained on netflix and other such services, [/INST] I encourage </s><s>[INST] you to go to express VPN dot com slash flex pod to get an extra three months free. This show is also brought to you by Master Class $180 a [/INST] year gets </s><s>[INST] you an all access pass to watch courses from the best people in the world, in the respective disciplines. The list of people is just insane from chris Hadfield who I really need to [/INST] talk to soon. </s><s>[INST] I've been trying to talk to him. We agreed to talk. We just haven't been able to make it happen, [/INST] but </s>

<s>[INST] that's a weird american warrior ethos that I've picked up, but also having the, the, the Russian, the city of brothers that don't see it, I don't see it as that they, they don't get draw, they think that there is a tension between the art of the martial art and the violence of the martial artist is a poetic way. I could put it, but they're not so fascinated with this dan gable dominating another human. They think the effortless, the effortlessness of the technique and your mastery of the art is exhibited in its effortlessness, how much you lose yourself in the moment and the timing that just the beauty of a timing, like there's much more like one example in judo, but also in wrestling, you can look at the foot sweep wrestlers in America and even judo players in America and much of the world don't admire the beauty of the foot sweep, but a well timed foot sweep, which is a way to sort of off balance to find the right timing to just effortlessly uh, changed the tape, turned the tables of uh, dominate your opponent is, is seen as the highest form of mastery and, and Russian wrestling and in case of judo, it's in japanese judo, it's interesting. I'm not sure, I'm not sure what what that tension is about. I think it actually takes me back to, I don't know if you listen to dan Carlin's hardcore history and Genghis khan, if you've ever, [/INST] I read a great, great book on Genghis khan. </s><s>[INST] I'm still, I'm still trying to adjust most of my life said genghis khan, but the right pronunciation is uh actually changes con. There's a tension there. We kind of think, I don't know, We, I kind of thought as Genghis khan is a ultra violent, a leader of ultra violent men. But another view, another way to see them is the people who warriors that valued extreme competence and mastery of the art of fighting with weapons, with bows, with horse riding, all that kind of stuff. And I'm not sure exactly where to place them on. My sort of thinking about violence in in our human history. [/INST] I think in a context of like a combat sports, I think there's a difference between an athlete winning a contest under a certain set of rules and a fighter winning a fight under those exact same rules, there's a different approach to it and I don't think one is any better than the other. Um like in M. M. A. I think a great example would be George ST pierre, Georgian. NPR's a tremendous, it's a tremendous athlete and he considers himself to be a martial artist first. He's trying to win an athletic competition. Like Nick Diaz is trying to bust your ass right? There's a different approach to it and yes, they've had different results at the highest level of competition, but it's difficult to attribute the difference in results just to their approach to the sport because they're different human beings with different abilities and different different physical attributes. Um The city of brothers have that luxury of being able to talk about the beauty of a perfectly timed slide by, right, there are other wrestlers that will never be able to pull that off and therefore they have to pursue other ways to defeat someone. And maybe it is that dan gable, breaking a man's spirit by out working him type thing, which it's beautiful in its own way. Uh, but we, we, we tend to self select the ways in which we're able to be successful and then kind of take a deep dive into that. </s><s>[INST] What do you think is more beautiful, brute force or effortless execution of technique that dominates another human? [/INST] I think it's a subjective thing based on what skills you perceive yourself to have. I'm never, I've never been a slick, super athletic, dexterous competitor in anything. And I've always been more of, uh, I've got to outwork, you've gotta out grind you, I gotta out mean you. And so because I've lived that, I tend to see the beauty in that more because I have a perceptual awareness that I don't have for the people who have the luxury of being very slick and athletic and, and, and using beautiful technique. Now that said, there's a phenomenal little video the other day, I sent to a friend of a compilation of foot sweeps by Liotta Makita in M. M. A. And they're so beautiful and they're so awesome and it's not that I don't have an appreciation for those, but I can't emulate those because I lack the physical ability to do that, whereas I, I at least have a chance to emulate some of the people who do it through grit and throughout working people. </s><s>[INST] But I would love to uh return to Jenkins con and get your thoughts about like I have so many mixed feelings about whether he is evil or not, Whether the violence that he brought to the world had ultimately the fact that it had maybe kind of like dan Carlin describes cleansed the landscape. It's like a reset for the world through violence had ultimately Mhm. A progressive effect on human civilization. Even though in the short term it led to massive, you could say suffering. I don't know what to make of that man. What what are your thoughts on genghis khan? [/INST] Um I think it's always difficult to look at a historical figure and their actions of their time through a modern day lens because it's very, it's easy for us to kind of, you know, in pune their achievements and the things that they did and say, oh well, you know what he did was wrong? Well, of course that can be true. But a lot of times we don't actually have any real good context or concept of the times they were living in and what really was deemed wrong and what really wasn't. We're looking at it through a very cushy modern lens that being said from what I've read about Genghis khan. Uh yeah, he was a violent dude. But also he gave you an option, He, when he, when he got to a village, he said, look, I'm going to, we're going to, you have a choice. You can come with us or you can run. And you know, he gave them an option to join his legion of fighters who he took very good care of. You know, he was the first military leader, uh, to pay his soldiers families when they died. And he did that based on the booty that they got when they raided a village, He took that money, he took his share and they divided that up amongst the soldiers and then the soldiers families. I think he also is credited with, uh, first like horseback mail routes or something like that, right is neither godfather of modern postal system. There's something, something </s>

<s>[INST] you're talking the metaphor that's popping up in my mind. And this is also something that bothers me with theoretical physics. The metaphor is there's some sense in which there's things called theories of everything. When you try to describe the basic laws of physics, how they interact together and once you do, you have a sense that you understand all of reality in a sense you do and that to me when you do that to me, is understanding the individual, like how the individual behaves in this world. But then you're saying that there, hey hey you're also forgetting chemistry biology, how all of that actually comes together. The stickiness, the stick stickiness of molecules and how they build different systems and they some systems can kill each other, Some systems can flourish, Some can make pancakes and bananas and some can make poison and all those kinds of things that we need to be able to, we need to consider the the full stack of things that are constructed from the fundamental basics. Um and uh I guess you're on, you're saying that no, you you're just like the theoretical physicists, it all starts at the bottom like if you need to preserve the fundamentals of reality, which is the individual, like the basic atom of human society is the individual to you. [/INST] So yes, so the basic unit, the basic model unit, the basic ethical unit in society is the individual. And of course we form groups and and you can't understand history unless you understand group formation and group motivation. And and I have a view about what kind of groups should be formed uh and and politically from a political perspective, voluntary ones, ones in which we we join when we want to join and we can leave when we want to leave and uh 1/2 that help us and clearly groups help us pursue whatever it is, a goal is ultimately So so in the pursuit of happiness, there are lots of groups that one wants to form, whether it's marriage, whether it's businesses, whether it's sports teams, whether it's lots of there are lots of different groups want to form. But the but the question is, what is the standard of well being? Is it the standard well being, some algorithm that maximizes the well being of a group, you know, some utilitarian function, you know, is it is it something that's inherent in the group that we can measure as goodness and to help with the individuals within as long as we can get that, that the group to function well, we don't really care about where the individuals are. So to me, the the the the goal of creating groups is the well being of the individual. And that's why it needs to be voluntary. And that's why there has to be a way out of of those, sometimes it's costly. It's not a cheap out. That's why you should really think about what groups you you you and this, you know, on an issue that's very controversial. Maybe we can discuss maybe not. This is why, to me, immigration is so important, right? Open immigration or free immigration is because that's another group that I would like people to be able to voluntarily choose both in and out. And and I'd like to see people be able to go and join that group that, you know, that they believe will will will allow for the pursuit of happiness. But let me say that, you know, that's a description of an idea what I'm just saying. I recognize that that's not the reality in which we live? I recognize that that's not the reality in which history, history, you know, recognizing that such as that the individual exists in a sense, philosophically is a massive achievement, right? You know, human beings however they evolved clearly. We started out in in a, in a tribal context, in which individual didn't matter. We followed the leader. The competition was for power, power over the group and and and and dictates how the group should work. Um You know, the history of human beings is a history of gaining knowledge and part of the knowledge is the value of an individual. And you can see that in in in religion, you can see that in philosophy, you can see that through the evolution and then, you know, we we evolved from tribes into nations and then empires and conflicts between nations and conflicts, empires. And we tried a lot of different things if you will. I don't think we always did it on purpose, but we kind of different philosophies, different sets of ideas drove us towards different collectives, different groupings and and different ways in which the structure And after, I don't know, 3000 years of kind of known history, uh history before that, but we don't know much about it. 3000 years of known history. You can sit back and evaluate and I think that's what that's what is done in the enlightenment and and you sit back and suddenly we can do it today. We can sit back and evaluate what promotes human flourishing and what doesn't and what do we mean by human flourishing? Human whose whose flourishing? Well, individual human beings. Now, since I don't believe in a zero sum world and the world is not zero sum, like we can see that it's it's empirically possible to show that the world is not a zero sum game. My fellow human flourishing doesn't come into your expanse. So I I you know, I can show that a system that promotes my flourishing will probably promote your flourishing as well and promotes the general welfare in that sense, because it promotes individuals flourishing. And we can we can look at all these examples of how we evolved and what leads to bloodshed and what doesn't and what promotes this ability to flourish as an individual. Again, an achievement, the idea of individual flourishing, and then we can think about how to create a political system around that a political system that recognizes and allows for the formation of groups, but just under the principle of voluntary. So you can't be forced to join a group, you you can't be caused into forming a group um other than the fact that you're born in a particular place in a particular, you know that in a sense, but that's not force that's there's a difference between metaphysics and between choices. </s>

<s>[INST] They kept talking about a closing window. Yeah, they really thought that was a closing window. I think it was the nature of that window is different uh, in that there was a kind of pacifism and it seems like if um war broke out, most nations in the vicinity would not be ready. He could buy [/INST] the people the </s><s>[INST] leaders that are in power. They weren't ready. So the timing is really right now. But I wonder how often that is the case with with leaders in war. That feels like the timing is now [/INST] the other commitment problem. The other shift that was happening that he wanted to avert that is kind of wrapped up with his ideology is this idea of a, of like a cultural and a demographic window of opportunity that if he wanted conditional on having these views of, you know, dramatic people and and and and a pure race and that that now is that he had had to strike now before any opportunity to sort of established that was possible. I think that's one, it's an incentive that requires his ideology as well. How </s><s>[INST] do you avoid it within this framework, would you say, is there ah that you kind of provide an explanation? But is there a way to avoid it? Is violence the way to avoid it? Because people kind of tried rational piece, peaceful kind of usual negotiation and that led to this war. Is that unique to this particular war? Let's say World War One, World War Two. So there's an extra pressure from Germany and both wars to to act. Okay, so we've highlighted that. Is there a way to alleviate that extra pressure to act? [/INST] Let me use World War as an example, suppose as many german generals said at that time, we have a window of opportunity before Russia. Where where we might not win a war with Russia? Like, so the probability that we can win a war is going to change a lot in the next decade or two, maybe even the next few years. And so if we are in a much better bargaining position now, both to not use violence, but if necessarily use violence because otherwise Russia is going to be extremely powerful in the future and they'll be able to use that power to change the bargaining with us and to like hold keep us down. And and the thing is is in principle, Russia could say, look, we don't want to get invaded right now. We know you could invade us. We know we're weak. We know we'll be strong in the future. We promise to like not wield our and abuse are or just merely just sort of take what we can get in the future when we're strong. We're going to restrain ourselves in future or we're going to hand over something that makes us powerful because that's the bargain that would make us all better off. And the reason political economists call it a commitment problem is because that's a commitment that would solve the problem and they can't make that commitment because there's nobody who will hold them accountable. So anything, any international legal architecture, any uh set of enforceable agreements, any U. N. Security council, any world government, any anything that would help you make that commitment is a solution. All right. If that's the core problem. Um, and so that's why, you know in magazine, you know, the the the lava Xena can do that. They can say listen, yes combo that strong today is going to be weak tomorrow. You have an incentive to eliminate this combo over here. But because they're going to be strong. But guess what? You're not going to do that and we're going to make sure we're going to promise that when these guys do get strong, we're going to restrain what they can do. Most of our constitutions in most stable countries have done precisely that right. There's a lot of complaining right now in the United States about the way that the constitution is apportioned power between states. Um, that was a deal. That was a commitment. The constitution was in the United States was a deal made to a bunch of states that knew they were going to be weak in future because of economic and demographic trends or guests they might be. And it said, listen, you cooperate And will and, and and will commit not to basically ignore your interests over the long run. And now, you know, 250 years later, we're still honoring those commitments. Um, it was part of the deal that that meant that there actually would be a union and so we do this all the time. So constitution is a good example of how um every country's constitution, especially country who's writing a constitution after a war that constitution and all of the other institutions are building are an attempt to like provide commitment to groups who are worried about future shifts in power. </s><s>[INST] And then does that help with avoid civil war. So could you speak to um, lessons you learned from civil wars? The american civil war and the others, [/INST] the Lebanon. one of the ways Lebanon had tried for a long time to um preserve the interests of minority groups, powerful minority groups who we're powerful at the time and knew that that the demographics were working against them would guarantee, you know, this ethnic religious group gets the presidency and this ethno religious group gets the Prime ministership in this ethnic and well, and a lot of, a lot of countries will apportion seats in the parliament to if the religious groups and that's an attempt to like give a group that's temporarily powerful some assurances that they're when they're weak in the future, that they'll still have a say, right, just like we portioned sheets in the Senate in a way that's not demographically representative, but it's like, unequal quote unquote in a sense to help people be confident that there won't be a tyranny of the majority. And now that just happens to have been like a really unstable arrangement in Lebanon, because eventually, like the de facto power on the ground just gets so out of line with this really rigid system of the presidency goes to this ethnic religious group and this prime ministership because that that it didn't last. Right. So, but you can think of every post conflict agreement and every constitution is like a little bit of humans, best effort to find an agreement that's going to protect the interests of a group that's temporarily has an interest in violence in order to to not be violent. </s>

<s>[INST] Yeah [/INST] right When you need to have to tell that story you have to understand that especially V. One you need to be able </s><s>[INST] to [/INST] articulate those those opinion based decisions and you need to own them and if you fail with some of them you didn't get it right. You then own them and fix them and move on. Right. Version one of the ipod wasn't perfect version one of the iphone wasn't perfect. We got a lot of opinion based decisions wrong. But as you go through because you got more data because V2 you had data on those original opinions and then you were able to then modulate off of that right and you still have new opinions because those are differentiators that you that we call differentiators that the things that you know move the move the product forward in its evolution but at the revolution state opinions, opinions, opinions, no data </s><s>[INST] and so you have this discussion you and steve in the stage and the whole team with opinions and there you have to be harsh, I [/INST] wouldn't say harsh but you have to be very determined, right. You know there there are two real opinion based decisions that happen on the iphone one was the keyboard, should we have a key hardboard keyboard or should we have a virtual keyboard? The Blackberry was the number one productivity messaging device of its time. It was called a crackberry for a reason because people loved it because it had it was easy to type and you know they could get their work done but when you're saying we're going to move from that, everyone's talking about that in the market and you say we're gonna move to a virtual keyboard and it's not going to work as well as the hardware keyboard. That's an opinion based decision, right? Because the data is telling you </s><s>[INST] yeah, [/INST] all the best sales are over here. </s><s>[INST] God that takes guts, [/INST] it takes guts but you have to look at it from a different point of view and this is how I learned to come to understand this because I had been building you know virtual keyboards before and I knew you know the goodness and the badness in them, right? But he was like look those are productivity devices were making it ours is born out of an entertainment device and productivity right? We need to show full screen videos, we are going to have apps, they're not apps, but our apps, the apple apps because there were no apps story yet are going to take over the whole screen. You want a full screen web browser, You don't want one that's like half of the device is just a keyboard. Maybe you don't need that keyboard in every instance. So we want that part of the screen to change based on the tool you may need at the time and maybe it's just full view. Right? So you have to go and understand it's a different type of device just because that's that and it's successful for that reason. The crackberry for the keyboard, that's not the only thing you're gonna do with this device because people only did messaging and maybe a few phone calls. This was gonna be so much more, it was going to be an entertainment web browsing device. So you wanted those tools to go away, but it wouldn't be as good as the hardware keyboard. So that's an opinion. Let me give you another opinion based decision that got turned around before it shipped steve said no simplot, I don't want any slots. We're going to make it very pure johnny was like of course no slots, johnny and we all looked around and go, that doesn't work. You can't do that. Well why does Very And then he would always and this was the magic of Stephen, like when you said no, that doesn't work. You go, well what does Verizon not have any sim slots right? They showed that you can do a mobile phone with those limbs and you're like okay here we go. And so a few days later we come back with you know and and so product marketing voice of the customer engineering. We all come back with all the data showing how many how many data networks and and mobile networks required sim cards versus did not and what the trends were. And we showed the data and that killed the or excuse me, brought back the sim slot on the original iphone because we're like because he was just like we're gonna tell a T and T to not use a sim right? We're gonna just tell him to do it differently, right? But we were like if we want this thing to go anywhere around the world, you want to put that friction in people who are gonna move from place to place. You know, they have different sims because of the prices and all that stuff. We had to show all of that data. And then that opinion based decision got turned into a data driven decision. And and the sim slot obviously showed up. So those are two very at the very same </s><s>[INST] time right [/INST] opinion can hold and so can data overrule opinion when data does exist for a V. One. </s><s>[INST] But the at the end of the day you don't know what the right answer is. So doing no uh sim card slot may have been the right decision. We won't know because maybe if if that was the decision then like many times throughout Apple's history you basically changed the tide of how technology is done. [/INST] Absolutely. You know, you </s><s>[INST] never know. [/INST] Apple started wifi. People don't understand why if I came out of there was no wifi in 2000 and one. Apple started wifi and then everyone else, you know, got onto on board. If you look at now where we're going, we're going to phones without sim slots because we have e sims right? And now the sim slots becoming legacy, legacy as a legacy ports, that legacy portal probably be gone by At six, maybe 10 years. They'll be gone. I'm pretty sure of that because it's so much easier for carriers, they don't have to have physical things to go out and write. So right now it's just the early days but it will happen and it will go its way, it'll fall away but it will take time. You just couldn't do it back then. </s>

<s>[INST] And that's ultimately the physical side. How well are you using for this entire mechanism, the energy available to you? And so given all the resilience to errors and all that kind of stuff, it seems that it's close to its maximum. And this paper aside, it does seem that life, obviously, that's the constraint we have on Earth, right? Is the amount of energy. So that's one way to define life. Well, the input is energy, and the output is what? I don't know. Self-replicating. Okay, let's go there. How do you personally define life? Do you have a favorite definition that you try to sneak up on? Is it possible to define life on Earth? [/INST] I don't know. It depends on what you are defining it for. If you're defining it for finding different life forms, then it probably needs to have some quantification in it, so that you can use it in whatever the mission that you're operating. </s><s>[INST] Do you mean like it's not binary? It's like a 7 out of 10. Life-like. [/INST] I don't know. I don't think that defining is that essential. I think it's a good exercise, but I'm not sure if we need to agree a universally defined way of understanding life. Because the definition itself seems to be ever evolving anyway, right? We have the NASA's definition. It has its minuses and pluses, but it seems to be doing its job. </s><s>[INST] But what are the different, if there is a line, and it's impossible or unproductive to define that line, nevertheless, we know it when we see it as one definition that the Supreme Court likes. That's kind of an important thing to think about when we look at life on other planets. How do we try to identify if a thing is living when we go to Mars, when we go to the different moons in our solar system, or we go outside our solar system to look for life [/INST] on other planets? It's unlikely to be a smoking gun event, right? It's not going to be, hey, I found this. You don't think so? I don't think so. Unless you find an elephant on </s><s>[INST] some exoplanet, then I can say, yeah, there's life here. You don't think so? No, but isn't there a dynamic nature to the thing? It moves. It has a membrane that looks like there's [/INST] stuff inside. It doesn't need to move. It doesn't need to move, right? I mean, look at plants. I mean, they grow, but there are plants that can be also pretty dormant. And arguably, they do everything that, as one of my favorite professors once said, that the plant does everything that a giraffe does without moving. So movement is not necessarily. </s><s>[INST] But on a certain time scale, the plant does move. It just moves slower. Yes. It moves pretty quickly. [/INST] I would say that, yes. It moves pretty quickly. I would say that it's hard to quantify this or even measure it, but life is definitely chemistry finding solutions, right? So it is chemistry exploring itself and maintaining this exploration for billions of years. </s><s>[INST] So, okay. So a planet is a bunch of chemistry and then you run it and say, all right, all right, figure out what cool stuff you can come up with. That's essentially what life is. Given a chemistry, what is the cool stuff I can come up with? [/INST] If that chemistry or the solutions that it embarks upon are maintained in a form of memory, right? So you don't just need to have the exploring chemical space, but you need to also maintain a memory of some of those solutions for over long periods of time. So that's the memory component makes it more living to me because chemistry can always sample, right? So chemistry is chemistry. But are you just constantly sampling or are you building on your former solutions and then maintaining a memory of those solutions over billions of years? Or at least that's what happened here. </s><s>[INST] Chemistry can't build life if it's always living in the moment. The physicists will be very upset with you. Okay. So memory could be a fundamental requirement for life. [/INST] I mean, life is obviously the chemistry and physics leading to biology. So this is not a disciplinary problem of one discipline, triumphing other discipline. But what you need to have is definitely, chemistry is everywhere, right? I tend to think you can be a chemist, you can study chemistry, you can study physics, you can study geology anywhere in the universe, but this is the only place you can study biology. This is the only place to be a biologist. That's it. So definitely something very fundamental happened here. And you cannot take biology out of the equation. If you want to understand how the vast chemistry space, how that general sequence space got narrowed down to what is available or what is used by life, you need to understand the rules of selection and that's when evolution and biology comes into play. Yeah, so the rules of natural selection operate to you on the level of biology? Rules? I don't know if there are any rules like that. It would be fascinating to find in terms of the biology's rules. That's very interesting and it's a very fascinating area of study now and probably we will hear more about that in the decades to come. But if you want to go from the broad to specific, you need to understand the rules of selection and that is going to come </s>

<s>[INST] But you have to select what you trained. So I don't [/INST] know. Yes, but like for how long like that at some point in your life that you might try something. So if you like it you know at some point in your life that you might okay let me try close guard. You might not like it now, maybe in two or three years from now, I </s><s>[INST] still don't like it. I kept trying it. Listen because listen, it's very difficult to get any respect and you just get it's hard to get to black belt and beyond in jiu jitsu at a respectable place and not have a good close guard. Close guard is [/INST] then don't do it. It's not necessarily </s><s>[INST] being a rebel. [/INST] No, no it's not it's I'll say because it's not it's not a position that you're under pressure that if you don't know you'll be in trouble, You're not gonna be in trouble not to know the close guard. You're just gonna go through for open guard. </s><s>[INST] I mean that there's not a problem. The main limitation is if you don't do a close guard a lot that you don't quite you don't get a full complete picture of understanding how to attack. Close guard when somebody puts you into a close guard when you're on top. So it's nice to know both sides. Just understand. [/INST] Yeah, but you can have a pretty good understanding of how to defend from the top and not having any bottom. </s><s>[INST] I mean some of it is also just like the length of legs and just the geometry of your body. Uh I'm sure marcello Garcia is a good close guard, [/INST] but </s><s>[INST] that's exactly, that's the point I'm trying to make. In theory, [/INST] theory, you can imagine </s><s>[INST] it, but for a hobbyist, I think it's interesting to think of that. Like is it possible to is it possible to focus on a small set of techniques that help you to [/INST] develop </s><s>[INST] still into a good [/INST] instead </s><s>[INST] of enjoying and still be [/INST] In the in the judicial ward 99%. I mean people that competes </s><s>[INST] even the people that compete [/INST] 1% max and you have high level competitors, I have no close close guard this. Okay, </s><s>[INST] thank you for me. You know, I think you would say that most people don't have a coast guard is such a, such a difficult position, understand for me, [/INST] Maybe one Day We'll Brainwash. </s><s>[INST] Yeah, good. I, I felt it's too easy to stall versus [/INST] attack </s><s>[INST] that that that was my main concern is like I want to be forced in every way to um to always be attacking to always be moving to always be and it felt like if I got really good, I've seen it happen with half guard too. It's like when people get really good at half guard, it just feels Stahle if you just look at the matches and so on. It's just slow things down to a thing that's not, you don't get reps on learning, you don't get action in interesting ways. So that was my worry that I'll get old and fat and just sitting in close guard all [/INST] day </s><s>[INST] holding onto the white boats trying to kill me because it's also, I mean that's the other thing for hobbyists and for everyone is to like when you first start, I [/INST] think </s><s>[INST] you have to relax in the face of the fact that you're just getting your asking now [/INST] stop. </s><s>[INST] That can also be really tough in the ego? I think probably the right way to see that is you're growing as a person. [/INST] You see that clearly when they are like in a bad position, let's say side mount or mount like a beginner, he will never relax on those positions. The moment they just say go, they like trying </s><s>[INST] to [/INST] people push you out and explode that there's no relaxation and work on the difference. It's like no, it's out and go until I have zero to give </s><s>[INST] until [/INST] I'm exhausted. My arms cannot </s><s>[INST] move. It's kind of fun to watch actually. What's the role of drilling did you do you like drilling? [/INST] I do not like drilling, but I'll tell you why I think fighting, it's mechanic, right? So it's, it's it's very important to drill a move until you learn the mechanic. Of course it's important if someone want to teach an arm lock, you want to practice that movement until you learn the mechanic of it. But the guy is not resisting. So it's easy to apply, right? So you apply as many times as you have to until you know, the mechanics of the moves until you can apply the mechanics, uh the moment that you know how to apply, there's no more point in drilling. Now, you have to practice. Now you have to practice with resistance. Of course. You're not going to practice with the guy fully resisting. The guy is better than you because he's not going to give you a chance to practice that move, but you have to practice with resistance. So what does drilling comes on? That is most of people, you know, they they flow drill and everything. I think whatever you do, you condition in your body to do something is that, you know, you do, you repeat the same move over over and over again. Your body's conditioning to apply that movement or the technique drilling is not realistic because the other person is not resisting. You know, the flow movement or whatever after you go beyond. When you already know that the the mechanics, the drilling with no resistance is not going to teach you anything, because you will never know how to apply the movement with resistance. So it's pointless to carry on drilling after you learn the mechanics </s><s>[INST] see but you're making it sound easy to learn the mechanics. I would argue [/INST] you can drill as many times. I'm not limiting how much you drill you drill as much as long as you have to. I mean it doesn't matter how long </s>

<s>[INST] It's a crypto asset that has monetary value. Let me just link on that Bitcoin is a digital asset that has monetary value. What is a digital asset? What is monetary? Why use those terms versus the words of money and currency? [/INST] Is there something interesting in that disambiguation of different terms? I'd call it a crypto asset network. The goal is to create a billion-dollar block of pure energy in cyberspace, one that I could then move with no friction at the speed of light. It's the equivalent to putting a million pounds in orbit. How do I actually launch something into orbit? How do I launch something into cyberspace such that it moves friction free? The solution is a decentralized proof of work network. Satoshi's solution was I'm going to establish a protocol running on a distributed set of computers that will maintain a constant supply of never more than 21 million Bitcoin subdividable by 100 million Satoshis, each transferable via transferring private keys. The innovation is to create that in an ethical, durable fashion. The ethical innovation is I want it to be property and not a security. A bushel of corn, an acre of land, a stack of lumber, and a bar of gold and a Bitcoin are all property. That means they're all commonly occurring elements in the world. You could call them commodities, but commodity is a little bit misleading, and I'll tell you why in a second. They're all distinguished by the fact that no one entity or person or government controls them. If you have a barrel of oil and you're in Ukraine versus Russia versus Saudi Arabia versus the US, you have a barrel of oil. It doesn't matter what the premier in Japan or the mayor of Miami Beach thinks about your barrel, they cannot wave their hand and make it not a barrel of oil or a quart of wood. </s><s>[INST] Property is just a naturally occurring element in the universe. Why use the word ethical? I'm sorry to interrupt occasionally. [/INST] Why ethical assigned to property? Because if it's a security, a security would be an example of a share of a stock or a crypto token controlled by a small team. In the event that something is a security because some small group or some identifiable group can control its nature, character, supply, then it really only becomes ethical to promote it or sell it pursuant to fair disclosures. I'll give you maybe a practical example. I'm the mayor of Chicago. I give a speech. My speech, I say, I think everybody in Chicago should own their own farm and have chicken, a chicken in the backyard and their own horse and an automobile. That's ethical. I give the same speech and I say, I think everybody in Chicago should buy Twitter stock, sell their house or sell their cash and buy Twitter stock. Is that ethical? Not really, but at that point, you've entered into a conflict of interest because what you're doing is you're promoting an asset which is substantially controlled by a small group of people, the board of directors or the CEO of the company. How would you feel if the president of the United States said, I really think Americans should all buy Apple stock, especially if you worked at Google? But he worked anywhere. He'd be like, why isn't he saying buy mine? A security is a proprietary asset in some way, shape, or form. The whole nature of securities law, it starts from this ancient idea, thou shalt not lie, cheat, or steal. If I'm going to sell you securities or I'm going to promote securities as a public figure or as an influencer or anybody else, if I create my own Yo-Yo coin or Mikey coin and then there's a million of them and I tell you that I think that it's a really good thing and Mikey coin will go up forever, everybody buys Mikey coin and then I give $10 million to you and don't tell the public, I've cheated them. Maybe if I have Mikey coin and I think there's only $2 million Mikey coin, and I swear to only $2 million, and then I get married, and I have three kids, and my third kid is in the hospital, and my kid's going to die, and I have this ethical reason to print 500,000 more Mikey coin or else people are going to die, and everybody tells me it's fine. I'm still abused the investor. It's an ethical challenge. If you look at ethics laws everywhere in the world, they all boil down to having a clause which says that if you're a public figure, you can't endorse a security. You can't endorse something that would cause you to have a conflict of interest. If you're a mayor, a governor, or a country, a public figure, an influencer, and you want to promote or promulgate or support something using any public influence, or funds, or resources you may have, it needs to be property. It can't be security. It goes beyond that, what the Chinese want to support an American company. As soon as you look at what's in the best interest of the human race, the civilization, you realize that if you want an ethical path forward, it needs to be based on common property, which is fair. The way you get to a common property is through an open permissionless protocol. If it's not open, if it's proprietary and I know what the code says and you don't know what the code says, that makes it a security. If it's permissioned, if you're not allowed on my network or if you can be censored or booted off my network, that also makes it a security. When I talk about property, the challenge here is how do I create something that's equivalent to a barrel of oil in cyberspace? That means it has to be a non-sovereign bearer instrument, open permissionless, not censorable. If I could do that, then I could deliver you 10,000 dematerialized barrels of oil and you would take settlement of them and you would know that you have possession of that property irregardless of the opinion of any politician or any company or anybody else in the world. That's a really critical characteristic. It's probably one of the fundamental things that makes Bitcoin special. Bitcoin isn't just a crypto asset network, it's easy to create a crypto asset network. It's very hard to create an ethical crypto asset network, because you have to create one without any government or corporation or investor exercising into influence to make </s>

<s>[INST] because you are better healed effectively. So I think we're speaking in sort of idealistic terms about the power of capitalism when it works well. Is there any aspects that you think that don't work well in a free market, in all the basic pragmatic truths that we were talking about? [/INST] Is there ways it can go wrong? So I would first argue that we have never seen an actual purely free market. Closest example would be kind of geopolitically, we have a free market and that governments are not necessarily governed, but they are premised on governing large groups of individuals. So this doesn't exactly- </s><s>[INST] You mean between governments? Yes. See, but isn't there still, so a free market, maybe you can correct me, is the free market still grounded in the ideas of property rights and all those kinds of things, right? So governments tend to also sometimes be violent towards each other. That's right. So they don't respect all the basic aspects of capitalism. [/INST] That's right, yes. That's right. That's right. So maybe another way to look at this is that gold is the original governor of government, actually. And this is the reason governments have abused and gone off of the gold standard. So historically, if you are a bank or a nation state and you produce more currency than your gold reserves can justify, then people, then gold will flow out of your bank or out of your country. So there's this natural check via the money that, via capitalistic money, which is gold, right? Gold was selected by the free market. It's not decreed by a government. It provided this natural check on government action. So my, I guess to get back to the original question is, we've never seen a pure free market because money has always been monopolized and coerced, frankly. So to try and answer what goes wrong with a free market is really difficult because we've never actually seen it. And I would define free market is one in which government only protects life, liberty, and property. And so that's a very minimal role in society. Again, just as network security for the economic trade network. And anything that, any government function that goes beyond those three core functions, which by the way, are pretty much the core tenants of morality as well. So government's just really intended to preserve natural law, if you will. Anything that goes beyond that is moves us closer to an unfree market. So every regulation, every act of coercion is actually a gradation closer to a purely unfree market, which would be a monopoly. So in terms of what I guess theoretically we could say goes wrong in the free market is that it's volatile. It's trading off, it's accepting short-term volatility in exchange for less long-run volatility. And this tends to be the way of nature, by the way. So if we could look at something like there's a region in North America called Baja, California, and it runs into the United States and it runs down into Mexico as well. So the same topology, but two different jurisdictions. In the US, we very heavily manage forest fires. We're trying to manage nature effectively. Whereas in Mexico, it's much more unregulated. It just, when wildfires spring up, they let them burn off. North America, when wildfires spring up, we're actually extinguishing them. So we're constantly trying to dampen the short-run volatility of these small brush fires. Whereas in Mexico, we just let them burn. We let nature do its thing. The consequence of this is that the wildfires still occur eventually, but they're much larger and much more devastating in North America where human intervention has occurred because it's dampening nature's natural corrective mechanism of clearing this underbrush with these more frequent and smaller fires at the cost of much larger fires. So again, we're delaying short-term volatility and exacerbating long-run volatility. And in Mexico, it's the opposite, right? Just these wildfires burn much smaller and more continuously over time. The further effect of that in North America is that the fires can get so big and so hot that it burns away the top soil. So it actually destroys the fertility of the soil itself. So the point of this is that human intervention, right? Even the intention behind North American authorities managing that forest fire is to create less destruction. That is the intention. But the intention is divergent from the outcome. So, and to let me speak, you would say that human intervention moves us from mediocristan into extremist stand. So mediocristan would be something much more like nature where for instance, you can't double your body weight in a day, probably can't even do it in a year, right? But in extremist stand, which is something much more information-based, you can double or send your net worth to zero in a single trade, in a single moment, right? So when we try and intervene with natural biological systems that have these feedback loops, we actually start to push the system more to behave more like an extremist stand system that has less short run volatility, </s>

<s>[INST] including within psychology and psychiatry. So it's nice to just think about this broad umbrella of narcissism and the levels to which it's benign or malignant, and then also separating into the different mechanisms, like interaction by interaction, which sometimes can be narcissistic, but broadly speaking, do you do everything through the lens of malignant narcissism that makes you a sociopath or a malignant narcissist? [/INST] Right, yes. And the thing I would add to that is the thought about culture, right, how does the cultures we're in, whether it's the culture of a household, right? The culture of a community, the culture of a nation or the world, how does that impact what unfolds in that person and then how does what unfolds </s><s>[INST] in that person impact that culture? Well that's the question is, What unfolds in that person? How does culture affect it? But how does your own psychological development unfold that? Because narcissism in leaders is the most impactful thing. Who are the most impactful individuals? What is the most impact of individual psychology we have? It's usually leaders of countries or leaders of major organizations and so on. And one of the things you mentioned with benign narcissism, that seems to be aligned with success. If you care about your own success, that's going to be, you're more likely to have narcissistic tendencies, I suppose. And so my question is, when you follow that threat of narcissism to become the leader of a country, now you have a lot of new interesting psychological complexities to deal with, like power. That old cliche, that power corrupts. Does that, is it possible for power to corrupt the human mind to where it pushes you farther and farther into malignant narcissism, into this destructive envy? What are your thoughts on power? [/INST] Like the effect of power in the human mind? Yeah, I think power is an accentuator, right? An intensifier, right? So I think it is true that there are people who can be, sort of in a gray area where there are malignant narcissistic tendencies and behaviors, but there are also ways in which that person can think outside of themselves and think in a broader way and think sort of kindly about others. And they're sort of trying to navigate whether they're aware or not, that they're trying to navigate between one and the other. And then the allure of power is, well, just look just exercise that power and you'll feel better, right? It'll show you, right? That you're good enough, look at the power you have and whatever may be going on in the person's mind, that then power, yes, can corrupt. Yes. I think that's why we have to have checks and balances, right? Because we don't, you know, we're all inscrutable to ourselves, let alone to others. So we must have checks and balances and we should always have them on ourselves as well as on others. We should want that for the health of ourselves and the world around us. So I think all of that is true, but there are also people who don't necessarily become corrupted by power, right? There can be an understanding and a grounding that there is a steward of power, right? A shepherd. I mean, the ways people describe utilizing power and utilizing it in a benign way that then fosters the healthy aspects of self, right? So like gratitude and humility, right? I mean, if we could add a healthy dose of gratitude and humility to everyone or to our society, there would be a sea change, right? But how do you feel gratitude? How do you feel humility? Those things are incompatible with narcissism, envy, right? With really the bad pole of things that we're talking about. And part of the reason I, you know, I'm so focused in my work and really what runs through all of my thoughts about life is the impact of trauma, right? Because trauma, you know, creates these false lessons and it walls us all from truth. And, you know, it starts to point towards the unhealthy ways of trying to feel better about ourselves. But we have the health in us too. We have those seeds of health too that can grow into being a steward of power and sharing power, being considerate and kind. And we see a lot of that in the world too, right? It's not all just the evil. You know, we see plenty of people who do good and who are generous of spirit. And, you know, we have both in us and it is, I think, you're talking about our culture and what the seeds that we sow and the climate that we set, including putting governors and boundaries around, like, how do we reign in or say that the more aggressive, the more envious or destructive is unacceptable, right? How do we foster the part that's kind and considerate and reflective and slow to judge of like, hey, let's learn a little bit more. Like, how do we foster that? And I think a lot of that comes back to early childhood education. I mean, I think we don't do nearly enough to protect children. And as a corollary to that, we don't do nearly enough to educate children. Let me say, I want to write a book. Second book I write is going to be everything I needed to know about life. I learned as a second year postgraduate psychiatry resident. I was like, why? You know, why then did I learn so much about unconscious motivation, about the impact of trauma, about how we can be envious and how we can act out, you know, even about how our emotions trump logic in us. Like, why don't we teach these things when we're young enough to understand like, why is that other kid bullying me? Right? Or why just because I'm a little bit bigger, do I want to go thump that other kid on the head? Like, what's going on? We don't do those things, you know, we're just, we're tripping ahead of ourselves, and we don't stop and think, how are we using our resources? How are we shepherding forward, the next generation, which, by the way, is the generation that's going to determine our fates too, right? As we get older, but we don't, you know, do that. I often think of like, in the Olympics, you know, you see like the great sprinters, right? And they've got to come out of the blocks perfectly, right? So if they come out of the blocks a little bit too fast, they're going to fall over, right? They're going to just fall forward. And I often see that in my head about us as humans and as a culture that we're rushing so far forward, we don't stop and say, wait, let's keep the basics here, the basic techniques of like, how are we navigating forward in life? Or do we just throw all those away? Because I can get some benefit by saying that you're bad, even though what's being leveled against you is wrong. Why? Do I take that? Or do I say, no, that there's something more important here that we want to shepherd forward in ourselves as a culture. And I think preventing childhood trauma and changing the ways that we educate children and adults would could again, make a sea change and maybe set us on a course towards, you know, even towards a greater likelihood of survival as </s>

<s>[INST] I think I interrupted you. Was there something else on cast refugee? [/INST] Just Casper FFG e is just kind of combines together these two schools. Um, and so basically it creates this system where If you have more than 50% that are honest then um, and you have a network synchrony than the thing kind of goes as a chain. But then if network synchrony fails, then kind of the last few blocks in the chain might um we kind of get replaced. But anything that was finalized by this kind of more asynchronous process gets uh, can't be reverted. And so you essentially get a kind of best of both worlds between those two models. </s><s>[INST] Okay, so I know what I'm doing to them. You're reading the cast refugee paper uh, apologize for the romanticized question. But what do you are some or the most beautiful idea in the world of ethereum? Just something surprising. Something beautiful. Something powerful. [/INST] Yeah. I mean I think the fact that money can just emerge out of a database if enough people believe in it, I think is definitely one of those things that's up there. I think one of the things that I really love about ethereum is also this concept of compose ability. So this is the idea that like if I build an application on top of ethereum then you can build an application that talks to my application and you don't even need my permission. You don't even need to talk to me. So one really fun example of this is there was this kind of game on a theory. I'm called crypto kitties. The just involvement of breeding digital cats and someone else created a game called crypto dragons where the way you play crypto dragons as you have a dragon and you have to feed it crypto kitties. And they just created the whole thing just like as an ethereum contract that you would send these uh these tokens that are defined by this other ethereum contract. And for the interoperability to happen like the projects and don't really need to like the teams don't really need to talk to each other. You just kind of interface with the existing program. </s><s>[INST] So arbitrarily compose a bill in this kind of way. So you have different uh groups that could be working. So you could see it scaling to just outside of dragons and kitties. It could be, you can build like entire ecosystems of software. [/INST] Yeah, especially in the decentralized finance space that's been popping up in the last two years. There has been a huge amount of really interesting things happened as a result of this </s><s>[INST] is a particular kind of like financial applications kind of thing. [/INST] Yeah, I mean there's like stable coins. So this is a kind of tokens retain value equal to $1 but they're kind of backed by crypto Cryptocurrency. Um then there's decentralized exchanges. So when as far as decentralized exchanges goes, there's this uh really interesting construction that has existed for about one and 1.5 years now. Called unit swap. So what unit swap is? It's a smart contract that holds balances of two tokens? We'll call them token and token B. And it maintains an invariance that the balance of token a multiplied by the balance of token B has to equal the same value. And so the way that you trade against the thing is basically like you have this kind of curve, you know like X times Y equals K. And yeah, before you trade it's at some point on the curve, after you traded just like pick some different any other points on the curve and then whatever the delta X. Is that steam out of a tokens? You provide whatever the delta Y. Is that steam out of the tokens you get or vice versa. And that's just and then kind of the slope at the current point on the curve and it is the price. Um and so that just is the whole thing and that just allows you to have this exchange for tokens and even if there's very few participants and the whole thing is just like so simple and it's just very easy to set up very easy to participate in and it just provides so much value to people. So </s><s>[INST] and uh the fundamental the distributed application infrastructure and allows that somehow. [/INST] Yes. So this is a smart contract meeting, this is all a computer program that's just running on ethereum </s><s>[INST] smart contracts to are just fascinating. [/INST] They are </s><s>[INST] Okay. Do you think Cryptocurrency may become the main currency in the world one day. So what do you think we're headed in terms of the role of currency, the structure type of currency in the world? [/INST] I definitely expect uh fiat currencies to continue to exist and continue to be strong and I definitely expect kind of fiat currencies to also digitize in their own way over the next couple of decades. </s><s>[INST] What's fiat currency by the way? [/INST] Oh just like things like us dollars and dollars and euros and yen and these other things </s><s>[INST] and there sort of backed by governments. [/INST] Yes, but I also expect and of cryptocurrencies to play this kind of important role in just making sure that people always have an alternative if fiat currency start breaking. So like if or if you're in, you know, some have a very high inflation place like Venezuela for example or if uh you're, my country just kind of gets cut off from like um cut off from other financial systems because of like something the banks do give any kind of, if there's even like some major trade disruption or something worse happens then like cryptocurrencies are the sort of thing that just because of the kind of global neutrality, they're just always there and you can keep using them. </s>

<s>[INST] Ridiculous question. Or perhaps not. If you look maybe 50 years from now or 100 years from that person born then what do you think? It's a good goal in terms of how long a person would live? Like what is the maximum longevity that we can achieve through the methods that we have today of uh or are developing some of the things we've been talking about. Yeah. And in terms of genetics, in terms of biology, what's is there a number? [/INST] Right, well so it changes all the time because technology is changing so quickly. I keep revising the number upward. But I would say that If you do the right things during your life and start at an early age, let's say 25, we don't want malnutrition, starvation, that's not what I'm talking about. But in your 20's start eating the kind of diets that I talked about, skipping meals In animals. That gives you an extra 20 2 30%. We don't know if that's true for humans and that would even 5% more would be a good a big deal for the planet. I think that we should all aim to at least reach a century. Um I'm a little bit behind. I was born too early to benefit the most from all of this discovery, Those of you who are in your 20's You should definitely aim to reach 100? I don't see why not consider this is really important. The average lifespan of a human that looks after themselves but doesn't pay attention Is about 80. Okay, Japan, that's the average age for a male bit higher. If you do the right things in your life, which is eight healthy food, don't ovary, don't become obese, do a bit of exercise, get good sleep and don't stress. That gives you on average 14 extra years that get you to 94. So get into 100. If you just focus on what I'm talking about, it's not a big deal. So what's the maximum while we know that One human made it 222 and a number of them make it into their teens. I think that's also the next level of of uh of where we can get to with the types of technologies that I'm talking about, medicines like I mentioned rapper Miss and there's one called Metformin, which is the diabetes drug, which I take That in combination with these lifestyle changes should get us beyond 100. How long can we ultimately live? We'll there's no maximum limit to human lifespan. Why can a well lived 300 years? But we cannot were basically the same structure. We just need to learn from them. So anyone who says, Oh you max max out at XI think is is full of it. There's nothing that I've seen that says biological organisms have to die. There are trees that live for thousands of years and their biochemistry is pretty close to ours. </s><s>[INST] What do you think it means to live for a very long time? Let's say if it's 200 years we're talking about for 1000 years. Yeah there's some some sense you could argue that there is immortal organisms already living on earth. Like there's bacteria. So there's certain the certain living organisms that in some fundamental way do not die because they keep replicating their genetic and they keep like cloning themselves. Is is it the same human? If we can somehow persist the human mind like copy clone certain aspects and just keep replacing body parts. Um Do you think that's another way to achieve immortality to achieve of a prolonged sort of increased longevity is to replace the parts that break easily and keep. Because actually from your theory of aging as a degradation of information. So an information theory view of aging like what what is the key information that makes a human? Can we persist that information and just replace the trivial parts. [/INST] Yeah I mean this short answer is yes we're already replacing body parts. But what makes us human is our brain. Everything else is suboptimal except our brain. Um The ability to replace actual neurons is really hard. All right. I think it might be easier to upload rather than replace neurons because they're so tight. It's such a network and just put Toby in the system. It's fraudulent this cat you you change everything once you get in there. The problem is um well I guess the solution let me go to the solution. That's more interesting. What we're learning is that if you reverse the age of nerve cells they look it looks like they get their memories back. All right. So the memories are not lost there. Just that the cells don't how to interpret them and function correctly. This is one of the things we're studying in my lab. If you take an old mouse that has learned something when it was young but forgotten. Does it get that back? And all evidence points to that being true. So, I'd rather go in and rejuvenate the brain as it sits rather than replace individual cells which would be really hard. </s><s>[INST] What do you think about like efforts like neuralink which basically you mentioned uploading are trying to figure out so creating brain computer interfaces that are trying to figure out how to communicate with the brain. But one of the features that is trying to record the human brain more and more accurately. Do you have hope for that? Two? Of course there's uh it will lead to us better understanding from a neuroscience perspective the human mind. But do you have hope for it increasing longevity in terms of how it's used. [/INST] I think that it can help with certain diseases. But I see at least within our lifetime, that's the best use of it is to be able to replace parts of the body that are not functioning such as the retina and other parts of the visual cortex back here. That that's going to be doable in terms of longevity. Maybe we could put something on the hypothalamus and start creating those hormones and get that back. Ultimately. I think it the best way to preserve the brain is going to be too recorded. But also I think it's going to require death unfortunately, to then do very detailed scans. Even if you have enough time and money, atomic microscopy and rebuild the brain from scratch. </s>

<s>[INST] it has a value floor and it won't collapse, engineering it. Is it possible for Cardano [/INST] and Doge going to work together somehow? Yeah, it'd be a lot of fun. I'm not adverse to the idea of cleaning up the code base, but legitimately whoever comes in, be two years or three years of work, because you have to do real stuff. And that code is like Litecoin circa 2012, 2013. </s><s>[INST] Well, the interesting thing about Elon, I've got to direct with him quite a bit, that combination of humor and extreme ambition, like in the face of impossible odds is something he does really well. And so I think that's the spirit of Dogecoin is fun and almost bold, ambitious innovation. And so I think you can't discount the power of that. [/INST] But where is the innovation? What's the- Well, this is step one. </s><s>[INST] What's going to Mars for Dogecoin. What's the- Well, this is step one. What's going to Mars for Dogecoin? Well, I mean, he came in the same way to rockets. He came in the same way to electric cars. It seemed impossible at first, but you step in and you solve the problems, first principles, one at a time. [/INST] But I'll challenge you a little bit on this because I think he had some trends that he was very smart to recognize. In the case of battery powered cars, he said, hang on, everybody has tablets and cell phones and these other things. And there's an incentive to make batteries better, faster, cheaper, and charge faster. And that's connected to mobile computing. So regardless if you want battery powered cars or not, every year you have billions of dollars of R&D being pushed to force this capacity to evolve. And he's just getting on the train and piggybacking on that. So that was a brilliant business acumen to recognize that. In the case of SpaceX, it was just obvious question. If every time you get on a plane, you have to throw the plane away, no one would fly. So reusability is like a fundamental thing that if you solve that, you've now opened space up to a complete new class of commercialization. I don't see the problem in Dogecoin. Cause if he was looking for it, then why not look at a real platform actually trying to solve a real problem? There's so many of them, he throw a rock, he can hit 15 of these guys and they'd all die to work with Elon Musk. </s><s>[INST] So many of them. Yeah, that's very interesting. I mean, so first I could continue pushing back on your intuition about electric cars and batteries and so on. I don't think it's more obvious in retrospect than it is at the time, I would say. Because I would agree with you on the battery's front, I wouldn't necessarily agree with the electric car. Because first of all, nobody started a successful car company for decades. [/INST] Well, but it was the loyalty. The EV4 guys or whatever it was with GM, they didn't want to give them back </s><s>[INST] when they had the lease program. So there's some basic intuition that there is some hunger here, but it's not obvious that you can do it successfully. And with relaunching rockets for cheap, that sounds good on paper, but to do it well, NASA is spending way more money for this. And the Russians were assholes not selling any rockets. So you have to do it all yourself from scratch. How do you build the team? How do you launch rockets when if you fail a few times, you're gonna go bankrupt? I mean, it's just business-wise, I would rather build an app, Angry Birds. Oh, yeah, yeah, he's got it. The launch rockets. [/INST] You gotta get right. Angry Birds. Oh, yeah, yeah, he's got it. The launch rockets. I'm crazy, he's got boulders for balls. </s><s>[INST] It's a good picture, good picture. Yeah, thank you for that. But I don't, maybe that's what, I mean, I think there's not enough first principle thinking on the cryptocurrency side. I think I agree with you on that. But there's some aspect to which the seriousness of the cryptocurrency world is paralyzing. So in some way, the innovation that you spoke to requires taking risks, requires not taking everything so seriously, like being afraid to take those bold risks in the space of ideas, not in the space of financials. So in that way, I think that's one pro for Bitcoin is there's room, it's hungry for innovation. But I think Cardano in that same way is hungry for innovation, just because you said with some more rigor and formulas and behind it. I mean, even Ethereum has a hunger for innovation. That's where Bitcoin is a little bit more, I would say conservative in terms of how much innovating they're willing to do, in terms of the incentives they built into the systems for the evolution of the cryptocurrency. But yeah, I mean, it's difficult to psychoanalyze why Dogecoin is the thing that excites Elon so much. But at the same time, there's some power to the fun. It sounds ridiculous to say. But the fun, this idea that the most entertaining outcome is the most likely, there could be something built into the physics of the universe that makes that true. Because the viral nature of fun has power. [/INST] Well, it's a neuroscience thing. We like dopamine, we like these chemicals in our brain. And when we have fun, we want more of it. So you tend to gravitate towards work activities and play activities that are enjoyable to you. And it is nice sometimes to kind of come in and troll an entire industry. Yeah. And I can imagine he's probably at the time of his life. It's the same as like taking Tesla private at 420. That kind of stuff. And it's one thing when you do it with friends, it's another thing when you do it with the whole industry and it hurts people financially. And also I understand he hates short sellers and there's some things there. And I can't speak for him because I only know the guy from a distance, it's just, I live in this space and I have to deal with the consequences and clean up the message. You feel the pain. Yeah, and this results in a regulatory event. It's like, I'm the guy who has to put on a suit and go to Washington and go and sit in a Senate inquiry and all this stuff. And he's the guy that just like laugh with his friends about how he crashed the crypto market and how easy it was to do it. And that's my only umbrage about this. On the other hand, if it brings a lot of cool, new, interesting things and people into the space, that's a net positive. And so, it's not universally bad or universally good. And I'd like to give people the benefit of the doubt. And so I'm really hopeful to see what comes about this recent surge of interest. And if Elon actually puts his money where his mouth is and takes some of his enormous engineering talent and capital and starts contributing and building something in the cryptocurrency space, be really cool to see that happen. </s>

<s>[INST] So fascinating that how much of a weird quirk of history is that um that religion would take hold in the country? Like does it have to do with the individual? It seems like if we re ran the 20th century 1000 times, when we get the 79 revolution resulting in um Islamic law, like less than, You know, 1% of the time it feels like or know what, which percentage would you put [/INST] it? I think it has something to do with the very complex nature Of how Iran evolved over a long period of time since the 16th century. That's why if I would for a moment talk about what I have written, I have written a book that's called Iran a modern history and it does not start in the 20th century. It starts in the 16th century because that's what I have argued that this complex process that at the end of today resulted in what we see around us today is something that was in making for a very long time </s><s>[INST] religion was a big part of it. She and the messiah complex, the [/INST] exactly </s><s>[INST] the longing for this great vision of a great nation that somehow is um the sublime nation that can only be fully sublime through through religion or [/INST] At the time it was thought that is through religion ever since then, it's disillusionment with that image or at least a process of disillusionment. The outcome of it is what we see today, basically that process of 40 years is a process of re adjusting to the realities of the world. That that great moment of romantic success of a revolution. Like most revolutions of course that is going to change Iran and bring this kind of a moment of greatest lead into this. Great disappointment is the movement of the great disappointment in a sense like most Messianic movements by the way music movements in general are always leading into great disappointments. But what I have here that perhaps should be added to it that yes, it was a peculiar reality of Iran as a society that had to experience this eventual encounter between religion and state. That's something to do with the nature of a visa. That's just one point that should be pointed out, most of Sunni Islam don't have that kind of I say most because there is something there. But Sunni Islam in general does not have that kind of an aspiration for the coming of a Messianic uh leader, she is um does she is, I mean, it's very shaping, particularly the way that it was set up in Iran was a religion that has always this element of expectation to it, for the coming of this Messianic leader. Of course, I mean between practices, all societies, look for Messianic leaders and just look around us. But </s><s>[INST] some societies more than others, there's certain culture, it might have to do with the romantic poetry that we mentioned over there. I mean, surely, I mean, not to draw too many parallels, but with the soviet union, there is romanticism to and I mean, I don't know that it does. Maybe idealism, [/INST] a sense of his savior, </s><s>[INST] who [/INST] would, who would bring you out of the misery that you are in and and always looking for for a third party to solve your issues. That's why probably this movement has a particular significance because it probably doesn't look for a messiah. Although I was talking to my brother who is a historian also and he was saying perhaps the messiah of this movement is that messiah mini, the 22 year old girl that was killed and martyred messiah, who is now leading a movement which no longer has that charismatic leadership with it. But yes, I would say that Iran has been the birthplace, if I might say that of Messianic aspirations going back to the ancient Zoroastrianism which is the really the whole system that you see in major religions or at least best so called Western religions. So abrahamic religions is parallel or perhaps influenced by Zoroastrianism, in which there is an idea of uh this word and the other word, there's hereafter, there is an idea of a judgment at the end of the time and there is a concept that there is a moment of justice that is going to come with the rise of a religious or a charismatic fear. So it's a very old phenomenon in Iran very old and it's time and again repeated itself in the course of its history, but never as powerfully As it happened in 1979 and never in the form of authority from within the religious establishment. It was always the dissent movements that were Kind of anti know me and they were against the authority of the religious establishment that changed in the 20th century. </s><s>[INST] But the Revolution 1979, that change is still with us today. What can we just linger on? Are there some practical games of power that occurred? You know, in the way that Stalin took power and held power in the early days? Is there something like this in terms of the establishment of the Revolutionary Guard and all those kinds of stuff? So the Messianic figure that has some support from the people, but uh does he have to crush his enemies in competition. [/INST] It certainly did probably not, certainly not as brutal in terms of the victims as you would see in soviet union under Stalin, who the bloodshed or the destruction of the population was far greater than what you would find in Iran of the Islamic republic, uncomfortable. Perhaps I would find a greater parallel with marty dung and particularly because china has a very strong Messianic tradition since the ancient times. So they have something and mao appeared as a kind of a Messianic figure there. I can see there is a parallel, but also you can see with any other authoritarian regime, with the Messianic fear at the head of it, that it destroys all the other forces. So during the course of the 1st 10 years of the Islamic revolution, it destroyed the liberal nationalist secular, it destroyed the guerilla movements, some of them Islamic, some of the marxists who turned into political parties or tendencies in the course of the post revolution 1979 they were completely destroyed and in a very brutal fashion. And uh the, their opposition even within the religious establishment, because it wasn't a uniform, there were many different tendencies, those that were opposed to the authority of Ayatollah Khomeini or not imam Khomeini meaning almost a sacred Religious figure above the level of a religious authority is a saint kind of a figure. She says she's um has this idea of imams, there were 11 of them, the 12 is hidden and would come back at the end of the time, this Messianic figure. So the title that was always used for them only in she's a never used for any other person. He is the first person in the revolution of 1979, 1st referred to as deputy of a mom. But the term deputy gradually disappeared and he became a mom for many. That's his official title. </s>

<s>[INST] what was to you involved in making games? Like how did you think of it? Was it from a graphics perspective? Like what shows up on screen, was it, how it makes you feel, was it about the story? Was that the text based stuff and the dialogue and the prompting? Like what, like um what does it mean to create a video game at that young age to you? [/INST] Well it was a way of experiencing things that I couldn't myself, so you know, if you're playing dungeons and dragons at the time to where you uh you really feel even pen and paper, these are like, like they feel somewhat in quotes reel to you as you're playing them, you're very invested in your character, what you're doing and then I love the games, the wizardry and ultima that we're able to bring that to a computer so I could, you know, do it on my own time was very, very real to me, I'd sit in my bedroom and then go to bed and think about it and then, oh no, I have to go to school, I want to come home and figure out how to how to do this problem in the game. And so whatever I was creating was something that I was excited about at the time, I made a Raiders of the Lost ark game. Um Yeah, so it was usually, you know, made a Miami Vice game, made it, grew the wanderer game, made a traveler game and made and but every time I was doing it, I wanted to figure out a new method on the Apple two of pulling it off graphically whether that was editing character sets to get graphics in different formats, or how can I enable the secret double high res mode? It had or just things like that, where it became kind of this limitless, what can I make this do? And I had some friends who were doing the same thing and then you get into who can impress each other, and I was kind of middle of the pack, I would say. Um and then, but again, this was the time where they're bringing computers into the school and the apples come into the school and the teachers are learning it because they have to teach the students, but then I was, I would say was part of a group of students that were like way past that. Um and it was very much of a self taught uh you know, how how do you make this thing dance? I'd </s><s>[INST] like to ask a strange question. So at that time, a lot of people consider you one of if not the greatest game designer creator of all time. You were middle of the pack then. Uh did you have a sense that this would be your life and you would also be creating, you know, the greatest games ever. [/INST] Not, not in the slightest. Um No, I don't think anybody, but I was very much like that was my dream at a, at that age, but you don't think that that's a job, you know, and the, as I got older, I was really going through college and I, that even the computer classes then weren't where I wanted them to be, so I still kind of doing my own stuff um and I ended up getting a business degree and then interviewing for some jobs like finance jobs, so well I guess I should do this to make money and I can keep doing this on the side. And I remember I actually got to like the final level of like this corporate finance job at circuit </s><s>[INST] city and [/INST] they turned me down and I was like suck them, I'm just gonna go make video games, so thank you. Circuit city. </s><s>[INST] Yeah, I think they went bankrupt actually, [/INST] they were based in Richmond, I was going </s><s>[INST] to school close to [/INST] there and so </s><s>[INST] so what, what's the origin story of you joining? But that's the soft works at the time. [/INST] So I had gotten Wayne Gretzky Hockey three for christmas from my girlfriend at the time, was now my wife. Um I was in college and I noticed that it was, you know in Rockville Maryland and oh, that's on my way home over christmas break back to William and mary where I went to college And I was at this point committed like this is what I want to do. So I'm just gonna drive by and knock on the door which is what I did. So I I drove by and knocked on the door as Martin Luther King Day 93 and someone came out and met me and said well maybe what? And I said well I'm in college, I'm talking about when I'm out of school like okay well contact us then and I will say I was I was I would contact them every once in a while. I did work for a small software company right um out of school down in that area of Williamsburg And still would contact Bethesda Arena had just come out. So they were in 94 arena had just come out and I loved it. I was in the sports games like the hockey stuff, they were doing a basketball, they did a basketball game. Yeah </s><s>[INST] I'm just looking at, they did a lot of, they did like six sports game six but that's the 10 games, six of them sports games and C double A Basketball, hockey league simulator [/INST] simulator. So it was really like sports gridiron which is like the first kind of physics based football game at the time. Um And there's a famous story with electronic arts trying to do Madden and then hiring Bethesda before my time to make Madden because they were struggling. Um When I started at Bethesda remember the owner had john Madden's Oakland Raiders </s>

<s>[INST] Thank you for listening to this conversation with Carl Kristen and thank you to our presenting sponsor Cash up. Please consider supporting the podcast by downloading cash up and using collects podcast. If you enjoy this podcast, subscribe on Youtube review with five stars on apple podcast, support on Patreon or simply connect with me on twitter at lex Friedman and now let me leave you with some words from Carl Freestone, Your arm moves because you predict it will and your motor system seeks to minimize prediction error. Thank you for listening. I hope to see you next time [/INST] </s>

<s>[INST] no [/INST] trash, no cigarettes, the ethic of like picking up your ship, there is like so intense, so it's not like the other festivals you go to where there's just trash everywhere, shit scattered everywhere. It's clean. People are picking up their stuff. People are like really being conscious of like not fucking up the plier. So I'm sorry, don't don't get a burner yapping about burning man that we won't stop till morning, </s><s>[INST] but there's a power there, but there is a power the culture propagating itself through to the stories that we tell each other and that holds up uh for Burning Man that like, it's clear that the culture has stayed strong throughout the [/INST] years, so </s><s>[INST] many people, so many really interesting people uh speak of Burning Man is like a sacred place they go to to to remind themselves about what's important. That's so interesting [/INST] and it is and it is, I mean it's like, you know, there are all these stories of like I love guru stories, I have a guru named baba never met him. He was Ram das guru, at least not in the flesh, but the story of the guru is if you're lucky you meet this being that and we're not talking about, you know, whatever the run of the mill, like charlatans out there. Like I know for sure that people are in the world right now who uh when you're around them, you the thing you're talking about the affirmation of the potential of humanity and also just an acceptance of yourself and you know, cultivated, like seeing someone who's cultivated love or compassion or whatever, but in this way that is yeah, I mean you would almost, you would rather meet that being than like a Ufo land in your backyard. It's like it is the ufo it's a person, but it's not, it's everybody and nobody and somehow they like end up conveying to you ideas that you may have heard a million times before, but somehow within the language itself is a transmission that permanently alters you. And so these people exist. I think you could argue that burning man the the total thing is a guru that a pilgrimage is involved to get there you like it's not easy to get there and when you get there it's gonna teach you something, it's gonna show you something it's going to and know what maybe some of the stuff it shows you might not be great but the community around you will like we'll hold you as you're like whatever the thing is that's coming out of it is coming out of you </s><s>[INST] and even the simplest activities, the simplest exchange of words have like just like with the gurus a profound impact somehow. Something about that place. [/INST] Yeah. Not to mention the insane synchronicities, like insane synchronous is here and I think like you know to get back to the notion of sentience as a byproduct of harmonized yet hyper complex system. Um I think synchronicities like like those kinds of systems are like lightning rods for synchronicity so crazy. Not just because you're high synchronicities happen that are that are impossible where you just have to deal with it and like you'll need something and within a few minutes someone's like oh here here you go. And </s><s>[INST] you mentioned but by the way Burning because of a psychedelic experience. Is that the strawberries or was it something else? What was what was the moment that was magical [/INST] it was D. M. T. Was strawberries. No no I was </s><s>[INST] yeah [/INST] I was like smoking D. M. T. And like I saw like if you're in the Midnight Gospel there these bovine creatures that have like a long neck and a lantern head. So like I saw one of those things and um and you know I thought it was funny and like ridiculous because you hear like all the Terence McKenna stories of the self transforming machine elves are all the purple or the magenta goddess. Everyone sees her. I'm like so this is what I get like fucking cow with a lantern head like that's where my brain is at and interacting with this molecule. So then like </s><s>[INST] I [/INST] look I look away </s><s>[INST] and [/INST] again this is D. M. T. So when I say look away, do I mean with my eyes shut? I look away, her eyes open, I look away. I think I shut so it sounds weird to say look away but however you wanna put it, that's what I did and I look back and it's still there only now it's you know because usually in like when you're having those kinds of visions they go away pretty quickly, things like moved like shambles ahead. Maybe a few steps just like a cow just like a cow. And then that was when the you know all the stories you hear about it like going through some kind of tuber. Some kind of light tunnel like a water slide made of light that's increasingly familiar. That's the wildest part of it is like, oh I know this place, not like I've seen this in like, you know on like bong stickers but like oh yeah, this is that place you go to, you just remember oh this place. And then it was like I was in some kind of, I don't know how to put it. A chamber, a technological chamber, some kind of super computer, some kind of nucleus that was technological and it was inviting. There was an invitation of like coming like come deeper into come deeper in and uh you can talk to whatever it is over there, you don't talk but there's a communication and I communicated but my friends, I don't, I love my friends, I guess I had some sense in that moment that it would mean complete obliteration or who knows what. And the response that it gave back was you can always go back there and that's when I opened my eyes and back totally, you know, and ever since then that that's caused me to revise my, my thinking on reincarnation, the idea that you die and you start as a baby and then live your life again. It goes right into what we were talking about, you know, that that maybe data, you know that the ship I saw in nitrous oxide don't feel dumb that my epiphanies are related to drugs but not all of them are a lot of them, But this notion of like, oh is it that we're imprinting into the medium of time space? Every thing we do and that that is a permanent imprint a frame that upon death can be accessed in the same way we can pull up pictures on our phone or computers and not only access but experienced as though in other words you could just jump in, you're still gonna have your memories, it's gonna give you a the illusion of having been a kid and got into that frame. But no, you just decided to go back there, nostalgia whatever. And uh yeah, </s>

<s>[INST] of one in a thousand. That's fascinating, because now I really need to listen to your conversation with Luke Cronin, who I believe you had, because he believes the opposite. Yes, I know. Yeah, I'm gonna have a debate with him. He believes that the way biology evolved on Earth could have evolved almost an infinite number of other ways. So like if you ran Earth over and over and over, you would keep getting life and it would be very different. So the fact that our particular life seems unique is just like, well, because every freaking life is gonna seem unique, but it'll be very different. It's not like, we shouldn't be asking the question of what's the likelihood of getting a human-like thing? Because that seems to be super special. It's more like, how easy is it to make anything that has the skills of a human? And I don't mean like something with thumbs, but achieving basically a technological civilization. And according to Lee, at least, it's trivial. [/INST] I know, I wanna have a debate with him. We fought a little bit. I'd love to debate him. I think it'd be a lot of fun. Because we debate with love. When I talk with Lee, I love him and he loves me, I think. I hope. But let me ask you a question. I asked this of him and Sarah on our clubhouse ones. So what do you think would happen the next day? Let's say we discover life, proximal century B. It looks just like slime mold, like you got on your brie cheese or whatever. We discover it. What would happen the next day? And they were like, oh, this would be transformative. And I'm not trying to be like total Cassandra about this, but I said, I don't think anything would happen. What are you talking about? This would be transformational. I'm like, I stipulate that life exists. Go down to like the river. I'm in San Diego, go down to the Pacific Ocean, scoop up a glass, you're gonna find life in there. And what are we doing? What are we doing to our earth? We're destroying it, callously. We're like pumping crap into there. Like we have this toxic waste spill a couple months ago in San Diego, I couldn't go to the beach. Let me take it a step further. You know how many people, I'm sorry that you do know, but how many people died in the 20th century killed? These are advanced civil, this isn't a slime mold. We kill, we maim, we harm, we hurt, we hate. I don't think anything would happen the next day. Then we go back to what we had. And I said, if that weren't proof enough, life has been discovered at least two or three times just in my professional career. Once in 1996, these Allenland Hills meteorites in Antarctica, like microbial respiration processes. Still, we don't know, it was a press conference held by Bill Clinton on the White House lawn that's featured in the movie Contact, repurpose for that movie. And then there's this phosphorous life, this toxic life in the pools of Mono Lake. Many, you know, extremophile, we don't give a crap. We continue to treat, so why are we thinking that like our salvation, from whence will our salvation come as the Bible says? Like, it's not gonna change how we are. It's not gonna magnify how I treat you or you treat me. And we're pretty knowledgeable people, you and I, </s><s>[INST] compared to, you know, lay people. Okay, that's interesting, that's a really interesting argument. I wonder if you're right, but my intuition is, I can maybe present a different argument that you can think about in the realm of things you care about even deeper, which is like, what happens once we figure out the origins of the universe? Like, how would that change your life? I would say there are certain discoveries that even in their very idea will change the fabric of society. I tend to see if there's definitive proof that there's life, and the more complex, [/INST] the more powerful that idea is elsewhere, </s><s>[INST] is that I'm not exactly sure how it will change society. Because it's such a slap in the face. It's such a humbling force, or maybe not. Or maybe it's a motivator to say, yeah, I don't know which force would take over. Maybe it would be governments with military start to think like, well, how do we kill it? If there's a lot of life out there, how do we create the defenses? How do we extract it? Or yeah, or mine it for benefits, all those. [/INST] I mean, I just see like there's 100 million literal counter examples of that. I mean, right now there's like 700 million kids in poverty, and like, how do we go about our life and just not deal with that? I mean, look, I put it aside. I eat hamburgers, and in the 100 years, I'll be canceled for being a carnivore or whatever. But, you know, so obviously to get through life, you have to make certain compromise. You're not gonna think about certain things. But I just think there is a sort of wish fulfillment. Like every time there's, why are we going to Mars and digging and flying this cool ass helicopter? We're looking for water. Like stipulate that water was there. Like, I believe there was water. I think we should invest again </s><s>[INST] and see what the geology was like. [/INST] But don't you think, so you're saying thing. Don't think you're gonna get meaning from it. That's all I'm saying. I'm not saying it's not worth doing. I'm just saying there's a wish fulfillment aspect that people will find meaning for life from science. </s>

<s>[INST] you know, I'm sorry to interrupt you, but uh you just made me realize that that's exactly right. So the visualizations you're creating while you're sometimes talking about abstractions, the actual visualization is a explicit low level example. Yes. So there's an actual, like in the code, you have to say what the what the vector is, what's the direction of the arrow? What's the magnitude? Uh Yeah, so that's you're going, the visualization itself is actually going to the bottom [/INST] I think, and I think that's very important. I also think about this a lot in writing scripts where even before you get to the visuals um the first instinct is to I don't know why. I just always do. I say the abstract thing. I say the general definition, the powerful thing and then I fill it in with examples later. Always it will be more compelling and easier to understand when you flip that. And instead you let someone's brain do the uh pattern recognition, you just show them a bunch of examples. The brain is going to feel a certain similarity between them. Then by the time you bring in the definition or by the time you bring in the formula, it's articulating a thing that's already in the brain that was built off of looking at a bunch of examples with a certain kind of similarity and what the formula does is articulate what that kind of similarity is rather than being a a high cognitive load set of symbols that needs to be populated with examples later on assuming someone still with you. </s><s>[INST] What is the most beautiful or awe inspiring idea you've come across in mathematics? [/INST] I don't know man, </s><s>[INST] maybe it's an idea of exploring your videos. Maybe not what like just gave you pause. [/INST] It's the most beautiful idea. </s><s>[INST] Small or big. So [/INST] I think often the things that are most beautiful are the ones that you have like a little bit of understanding of but certainly not an entire understanding. It's a little bit of that mystery. That is what makes it beautiful. </s><s>[INST] Almost the moment of the discovery for you personally, almost just that leap of uh ha moment. [/INST] So something that really caught my eye, I remember when I was little there were these like um I think the series was called wooden books or something. These tiny little books that would have just a very short description of something on the left and then a picture on the right. I don't know who they're meant for, but maybe it's like loosely Children or something like that, but it can't just be Children because of some of the things that was describing on the last page of one of them. Some were tiny in there was this little formula that on the left hand had a sum over all of the natural numbers. It's like 1/1 to the S plus 1/2 to the X. Plus 1/3 to the S. On and onto an infinity. Then on the other side had a product over all of the primes. And it was a certain thing had to do with all the primes. And like any good young math enthusiast, I'd probably been indoctrinated with how chaotic and confusing the primes are which they are. And seeing this equation where on one side you have something that's as understandable as you could possibly get the counting numbers. And on the other side is all the prime numbers. It was like </s><s>[INST] this. Whoa! [/INST] They're related like this. There's a there's a simple description that includes like all the primes getting wrapped together like this. This is like the oiler product for zeta function. As I later found out the equation itself essentially encodes the fundamental theorem of arithmetic that every number can be expressed as a unique set of primes to me still, there's I mean I certainly don't understand this equation or this function all that well, the more I learned about it, the prettier it is the idea that you can this is sort of what gets you representations of primes. Not in terms of crimes themselves, but in terms of another set of numbers that are like the non trivial zeros of the zeta function. And again, I'm very kind of in over my head in a lot of ways as I try to get to understand it. But the more I do it always leaves enough mystery that it remains very beautiful to me. So </s><s>[INST] whenever there's a little bit of mystery just outside of the understanding that and by the way, the process of learning more about it, how does that come about? Just your own thought? Or are you reading reading [/INST] or is the </s><s>[INST] process of visualization itself revealing more to you [/INST] visuals help? I mean in one time when I was just trying to understand like analytic continuation and playing around with visualizing complex functions. This is what led to a video about this function. It's titled something like visualizing the Riemann zeta function. It's one that came about because I was programming and tried to see what a certain thing looked like. And then I looked at it and like, whoa, that's elucidating. And then I decided to make a video about it. Um but I mean, you try to get your hands on as much reading as you can. You you know, in this case, I think if anyone wants to start to understand it, if they have like a a math background of something like they studied some in college or something like that. Um like the Princeton Companion to Math has a really good article on analytic number theory. And that itself has a whole bunch of references and you know, anything has more references and it gives you this like tree to start flying through. And like, you know, you try to understand, try to understand things visually as I go, that's not always possible. But it's very helpful when it does you recognize when there's common themes like in this case cousins of the fourier </s>

<s>[INST] the hope is that the wisdom is in the founding documents and the imperfect. But why is founding documents of that celebrated freedom that celebrated all the ideas sort of having a lot of nukes, having a lot of airplanes and tanks. It's not that's not that's not important. And the hope is whatever we're doing here with this quote, greatest country on earth, that we preserve the ideas and help them flourish. He said, [/INST] well, that's what I mean. So we could get so if you go back to the wu tang, I'm saying that's what we're striving for. We're starting for that, you know, </s><s>[INST] but you started unpredictable and just like, yeah, but like got deep pretty quick. I got to talk to you about Bruce lee, who's Bruce lee to you? Who is he to the world, What ideas of his or interesting to you? Like what you know, you talk about like Hendrix and music, Bruce lee is that in martial arts, he just seems to have changed the game. [/INST] You know, I went, I guess, I don't know what the word bold is the right word to say, but I went as bold as the same that he was a minor profit. And I got that concept from the holy Koran, what it says that we send profits to every nation, every village. We don't let nobody not here the world in some form because it won't be fair. And so if the law is most of all, even a man whose death has to somehow get a sign, I don't know if moses saw the burning bush was nobody else to talk to us. So we had to talk to the bush. I don't know. It could have been the bush, this one too. But point being made, it says that they are mining profits. And I see Bruce lee as one of them. Uh, because what he brought to the world through it through martial arts, uh, it was a whole shift in the dynamic of thinking. You know, and that happens when certain certain entities are born, but he didn't do it only in uh, in the physical sense. He was also full of philosophy and in the same process. Uh, and he was also striving to be the best of himself. So you got three things going on. Uh, study Bruce lee multiple times. At 1st, of course. Um, when I saw my first kung fu movie, it was the fake, it was, it wasn't really Bruce lee, it was a few greenhorn Eclipse cut together. And then I saw a black samurai. Then my following kung fu movies was like Fearless Fighters, uh, the ghostly Face, uh, you know the first double cable basically and feel his fighters, the lady put the little kid on her back and flew across the ocean across the lake. Right? So Bruce wasn't doing that. And then I went on to five deadly venoms and spear men and 36 chambers. And these movies are beautiful and yet they're all heightened Bruce, their heightened beyond doable. We're not gonna </s><s>[INST] like surreal, they play with the world, that's not of this world, [/INST] Bruce played with this world. So when somebody so many, I first saw Bruce honestly didn't think he was as good as these guys. He can't fly, </s><s>[INST] he's [/INST] not flying in the movies, right? Um, but then when I saw the first one I saw was the big boss, was they re titled Fist of Fury. But then when I saw chinese connection, which is uh, the real fast food, right? I saw something different there and I, he got Enamored and then of course, enter the dragon, right? Just really complete. That's why my first album is Enter the Wu Tang 36 Chambers of Charlotte. So it's enter the dragon, the 36 put together because those are the two epidermis. So what happened is you know that's young me then teenage me studies him again and I realized, wow look at look at his physicality, look out how he's really, he's moving for real and then I studied him again, wow, look at what he's saying. Then I studied him again while look at what he stands for. </s><s>[INST] Which do you like in the realm of martial arts? The reel or the surreal or the dance between the two? Yeah, [/INST] I like to dance between the two because a move, I mean a movie to me is to entertain you. So I'm cool with Obi wan Kenobi disappearing out of the cloak invader strikes him down and then I'm like yo what happened? He's like one luke run. I'm cool with that right? Because that's the imagination and the imagination gets stimulated to the point to as things that we saw imagined by our artists, we strive to create in our real world. Thus star trek to me is just a precursor to our cell phones. So for me uh I like to mix the two. </s><s>[INST] Yeah, it's funny how the like science fiction, like pushing into the impossible actually makes it realize eventually. Yeah, we humans like once we see idea on screen no matter how wild it is, we're trying to make it. Yeah we're trying to make it. It's only a young kid that gets inspired and watch that would be like I'm going to build that. [/INST] Exactly. So I don't know who is going to come with the back to the future time machine. But uh, do you have any classmates that you think that time </s><s>[INST] machine? I thought you were going to back to the future? Like the what is the hoverboard or like [/INST] we were there at least somebody they got, they see the one in the water, </s><s>[INST] you know the, [/INST] you know the surf hover? It's dope. It's dope. It actually it actually if you were back to the Future fan, you feel like you made it to </s>

<s>[INST] This show is also sponsored by simply safe home security company. It was designed by chad and Eleanor Lawrence to be simple and effective. I don't know why they asked me to mention the names of the creators But I think the point is that there's real human beings that really care behind this company and that actually when it comes to security systems is very important. It is indeed simple. Takes 30 minutes to set up. You can customize the system for your needs on simply safe dot com slash lex. I haven't set up in my place and I love it. I do try to live life, not paranoid because I think minimization of stress versus maximization of safety is the more effective way to live life for me. But I do make sure I make intelligent decisions of how to protect myself, how to protect my data, how to protect my physical safety. Especially as my platform grows. It gets trickier and trickier because I do want to keep myself open and fragile. All the people I meet on the street that recognized me from the podcasts are just incredible people. I really enjoy those conversations so I never want to close myself off from that. Anyway. Go to simply safe dot com slash likes to customize your system and get a free security camera plus a 60 day risk free trial. Again, that's simply safe dot com slash lex This show is also sponsored by ground news and a political news website that helps me get all perspectives on the story and analyze my blind spots politically. They draw from 50,000 outlets across the world and across the political spectrum. The point is to see every side and come to your own conclusions. I think this approach is the future of news. That's why I'm a big supporter of ground news. That's why I took him on as a sponsor. I see my membership there as support for the ideal of what they stand for, not just this particular company. Hopefully you can see it the same way it's been said before, but probably can't be said enough is that Clickbait journalism is truly damaging our ability as a society to have difficult long form conversations as we strive towards the truth. Truth is not an easy thing to arrive at. It requires understanding context that requires understanding history requires understanding lived experiences to different people to be empathetic to be data driven to be rigorous with your analysis, all of those things. And click bait journalism is completely antithetical to that. So that's why I think you should support ground news and other companies of that kind anyway, try them out by signing up at ground dot news slash lex. It's inexpensive so it's definitely worth it. But like I said, it's also a good way to support all the different efforts that try to fix our current state of journalism and the media go to ground that newsflash legs to sign up and show your support this is a lex Friedman podcast and here is my conversation with Jacko willing. Is it tragic or beautiful to you that some of the closest bonds that are formed between people are through war often? [/INST] I think it's both both tragic and beautiful and for the obvious reasons. </s><s>[INST] Yeah. What are the obvious reasons why is it so obvious? [/INST] Well, it's tragic because a lot of people die and it's beautiful because you form bonds with people that are very difficult to break once you've been through them. </s><s>[INST] What is it about the trauma of war that makes bonds difficult to break? [/INST] Because what you realize when you're in the war is that the people that are next to you, you rely on them and they're relying on you to survive and without them you will not survive. And when you realize that you need to work together as a team two to live that forms a very strong bond. </s><s>[INST] And there's nothing like that team outside of the realm of war. [/INST] I don't know because I've there's a lot of things that I haven't experienced in my life, but I think the pressure and the consequences of war, there could be similar situations in survival scenarios in various atrocities where people need to work together in order to survive. And I think you could probably get something that was similar. </s><s>[INST] There's a very particular nature to the kind of war that World War Two was especially for the soviet union where it didn't just influence the lives of people have created culture, the music, the poetry, the literature, it's in the it's in the way people think it's in the way people see the world, in the way they talk even still to this day. And of course I was talking about the directly relationship between two soldiers, but there's something about the depth of human connection that results from the almost like reverberations of war. Like generations later you're still close to other humans. Your there's a coldness towards other humans in Russia. But once you open up its depth, you seek depth of connection versus like breath of uh career kind of thinking how can I make friends with this? So I can move into this direction. What can this person benefit me? You instead you seek a depth of human connection and appreciation that that brings a lot and maybe I'm romanticizing war here, but it feels like that's inextricably connected to World War Two. For Russians, Does that resonate at all? [/INST] So if you look at military training, what they do is they take people in the military, from the civilian world, they bring them into the military and they put them through boot camp, which is the stereotypical thing that you see on tv, you're going to get yelled out, you're going to get screamed at, you're going to get you're gonna get put in the mud and you're going to need to be made to do hard things together. And what does that do with those civilians while it gives them a common background, it gives them a common suffering that they've been through together and they form some sort of connection. Some sort of bond now to make that bond a little bit stronger after you get done with boot camp they send you to advanced infantry school and you suffer some more together. And when you suffer more together now you're in a smaller group too because now it's infantry, it's not supplied people anymore or logisticians, it's strictly people that are going to fight their infantrymen. So they go through school together and now they get a little bit tighter get done with that and maybe you go to an airborne division. So you go to airborne school and now you all overcome this fear of jumping out of an airplane together and you celebrate surviving that. Then maybe you get done with that. And now you're at an airborne division now you're an even tighter group because you've suffered together. What comes next is special forces training or ranger training and what they do is they put you in these situations where you're gonna suffer together and you're gonna build these bonds because as I said earlier, you have to rely on each other to survive. And by the way not everyone does not everyone makes it through this training. So you have these memories of people that didn't make it, you share that connection as well and you can keep going down this road until you go into combat with a military unit and military units that go through combat. I haven't even tighter bond. And the harder the combat that they go through, the tighter the bond is going to be. So I think when you talk about what the soviet union went through in World War Two, there was a shared suffering to survive. And so the entire nation has that common thread and that's probably the thing that you sense or feel when you refer back to the bond that resonates all the way back to World War Two. </s>

<s>[INST] can you like get him back by acknowledging like the elephant in the room that like [/INST] that helps, but they're still going to go and it was funny when he made fun of it. But he sucks. He's still son still sucks. That's the worst part you're going, No, this is good. You guys just don't like me just because you don't like, it doesn't mean I'm bad. </s><s>[INST] Yeah, I like going to open mics a lot. Just, just listening because first of all, I think the audience and open mic at least the ones I've been to is mostly guess other comedians or like at least people who don't seem to want to laugh at anything. And so I just love it because it's human nature and perseverance that it's best that here's comedians like clearly, uh this is mostly in Austin they have a dream, like why would you get up there? [/INST] Maybe some weird, you know, New Year's resolution bullshit, but for the most part, it's people who want to </s><s>[INST] be a comedian, like a lot of the open makers are people who have clearly have done this for quite a long time, like at least a year or two, maybe five years, and they're often not very funny and um The just bombing in front of an audience of like 20 where they're just sitting there, like almost like mocking them with their eyes or maybe, and I don't know, and they still push through, they still they still like as if they're doing an arena and everybody's laughing, they still, they still got that energy trying almost like to an audience that doesn't exist like an audience of their dreams because I guess you have to do that to keep the energy of the act going and it's just so beautiful to watch them try it. It's uh and also the, what happens, open mic, I don't know, five minutes, whatever they do, they, you know, walk off and that walk back, you know, offstage and like you can't, who do they look at like, what do you look, do you make eye contact with people, do you, [/INST] you look at your phone, you look at your feet, you just zone out, you kind of, you kind of go white, you know, you just hear white noise and go out. It's, it's tough, But you got it. You need a little, a little delusion to be a comedian to get into it. It takes a little bit of delusion. Like, you think you can do this, you know, you got 10 years ahead of you of hell and you're up for this and you know, most comics, we see a horrible crowd and we see our friend bomb and we go, yeah, he's bombing. But I'll get him, I'll get them and then you don't get them. But that's, that's human nature too, is like, I don't like him, but they'll like me and you need a little of that to keep going as a comedian, but you don't want too much delusion because then you're a psycho, but you need a little, </s><s>[INST] well, the cycle could be good for a comedy. [/INST] That's true too. Like </s><s>[INST] I mentioned to you offline, um, I talked to Ellen and we talked about doing stand up that he's thinking maybe do a few minutes of stand [/INST] up saying if you need a coach, Ellen, I got you. </s><s>[INST] Uh, well maybe you should move to Austin to coach in full time, hopefully [/INST] can fly me in. </s><s>[INST] So what, what advice would you give to somebody who, um, who wants to try and do five minutes, like the early steps of trying to go to an open mic and say something funny. [/INST] Well that's the irony of comedy is, I don't know if it's irony, but it's like the beginning is the hardest part. Usually the beginning is easy part. Hey, I'm playing this level of Mario at startup, jump over one Koopa troop or whatever and then the end is like jesus christ, I got 30 guys coming at me. Comedy is the opposite. The beginning is like, it's a gauntlet, it's just obstacles and it's like you said open mics, you, I watched these famous comedians on netflix and you go, this would all bomb at an open mic there, killing in, you know, radio city, this would bomb on an open mic. That's the weird part. So it's almost like you have to go through hell just to get to the promised land and uh, I would say rehearse the shift out of it because you're gonna get frazzled up there. Everybody thinks, oh, this is good material. But you also forget about the other part of delivering it, Having confidence, being likable, having timing, having a cadence, figuring out who you are, figure out what the audience thinks you are or how they perceive you because you can go up there and say all this, but they go, that's the guy, he's clearly gay, Why is he acting like he's not gay? You know, that's all there now they're not listening to the jokes. So like you got to know how you look and uh it's just repetition, repetition and bombing is not failure. That's what you gotta remember. I mean look, if you, if you do a killer hour and then you take it to netflix and bomb your fucked up. But bombing is not failure, it's just data, it's going okay. I gotta re re tool that that didn't work something wrong there. They missed a word there. So you've got to treat the, the act almost like like, like ingredients in a, in a, in a cooking in a dish, you know like oh that I put too many eggs in, take an egg out. You gotta treat it like that and look when you pull it a bad cake out of an oven, you go, I funked up but it doesn't hurt your feelings, but when you bomb and funk up it hurts your feelings so you gotta factor that into your feelings gonna be hurt and just almost be a robot and just keep going towards that open mic. You know how scary it open mic is, bombing sucks but bombing in front of other comedians is way worse because they know what just happened and they could have saved you and they didn't so it's way worse </s>

<s>[INST] see you're still carrying the burdens of that decision. But yeah, you know, sleep is one of those things that's making me, I wonder about the variability between humans a little bit and how science is often focused on, like it's not often focused on high performers in a particular way and it's looking at the aggregate versus the individual cases, for example, like for me, I don't know what the exact hours are, but like power naps are incredible. Mhm. I tend to look at the metric of stress and happiness and joy and try to optimize those, so decreasing stress increasing happiness and using sleep as just one of the tools to do that because like hitting the 56789 hour mark or whatever the correct mark is, I find that to be stress inducing for me versus stress relieving, like thinking about that. I feel best if I sleep sometimes for eight hours sometimes for four hours and the power nap and as long as they have a stupid private usually smile in my face. That's what I'm doing. Good. As opposed to getting a perfect amount of sleep according to whatever the latest blog post is. And I also pull all nighters still, I also think there's something about the body, like as long as you do it regularly, it's not as stress inducing, like, you know what, you know what it is? The reason I pull all nighters isn't for like I'm playing Diablo three or something is because I'm doing something I'm truly passionate. Well I'm also lovely against, but I'm doing something I'm truly passionate about and it's almost like there's the jackal willing feeling of when I'm up at seven a.m. And I haven't slept all night and still, I'm working on it. There's a kind of a celebration of the human spirit that I really enjoy it. Like uh and that's happiness and to sort of then, and I usually don't tell that kind of stuff to people because their first statement will be like, you should get more sleep. It's like, no, I'm doing stuff. I love, you should get more love in your life, bro. That's right. So, but that said in aggregate, when you look at the full span of life, it's probably you should be getting a consistent amount of sleep and it seems like it's in that 78 hour range. [/INST] Yeah. But it's similar to food. It's the quality, not the quantity, right? And when you get it, so I look at my my data pretty often and what makes a difference to me is not the amount of hours, but the quality, the depth and the deep sleep is what, what we'll do it. So if I have a lot of alcohol before going to sleep and I can see my heart rate being different. But what really kills me is that I don't get a lot of that deep sleep. And I wake up barely remembering stuff so that like you say, if you're happy and contented and you're not gonna have these quarters old chemicals going through your body, you are more naturally get into that deep state. And even if you just get four hours way better than eight hours of none of that. </s><s>[INST] Yeah, that's beautiful. And some of that could be genetic for me, I just, I fall asleep like this. And if you want me to fall asleep right now. All I can do it. It's no um I have no problem with the combined with coffee, I just had to energy drinks. I could probably uh sleep so that I don't know if that's genetics or it's kind of um I don't know what it is or maybe that I don't have kids that I'm single. So I don't have, I'm almost listening to some kind of biological signal versus societal signal and what I'm supposed to go to sleep. So I just go to sleep. Whatever. I feel like going to sleep. Well, that's [/INST] because you're self employed self. Most people don't have that luxury, but we're lucky the two of us that we can make our own hours. Yeah. But yeah, it's super important. And those people who have shift work, I mean, they really need to change the way that works because they're literally killing those people. </s><s>[INST] Is there something you can say about the the mind and stress uh in terms of effect on longevity, sort of uh because I don't know if you think about it this way, but when you talk about the biological machine, it's always these mechanisms that don't are not necessarily directly connected to the brain or the operation of the brain. Like what's the role about stress and happiness and uh yeah, the sort of higher cognitive things going on in the brain on longevity, [/INST] right, well, that's a great point that the brain is the center for longevity actually. We do know that first when, when I'm stressed, I can see mentally stressed, then I can see it in my body. Heart rate hormones. It's clear. That's not true surprise. So you've got to work on your brain first and foremost, if you are totally freaked out agitated all the time, you will live shorter, I'm certain of it. I keep fish a big accordion guy and you can see the difference between the fish that's having a good time and dominant and one that gets picked on, it just looks like crap. You don't want to be that the little fish getting picked on if you can help it. So I used to be extremely stressed as a kid. I was a perfectionist, very shy, always worried about being a failure if I didn't get an A plus crying in my bedroom, that kind of sad existence. I got into my twenties then in my thirties and realized that's not the way to live. So I've worked very hard to get to this point where I almost never get stressed, never, there's nothing that I've never gotten angry in my lab. I got 20 kids. Sometimes it's like a most of the time it's like a kindergarten, I haven't lost my temper very calm, but that's intentional and I don't worry about stuff millions of dollars, billions of dollars at stake sometimes. Keep it cool. It's only life. We're all headed to the same place anyway. Don't worry about it. But the to answer your question I think in a better way if you manipulate the brain of an animal, I'll give an example if we turn on this certain gene that I mentioned. Someone we good friend of mine at wash U. And um I did this they up relate that operated that gene just in the neurons of the animal. It lived longer so that's sufficient to extend lifespan. We also know that you can manipulate the part of the brain uh called the hypothalamus which leeches a lot of chemicals into the body uh and proteins most of which we don't know yet. But just changing the inflammation of that little organ um or part of the brain is sufficient to make animals live longer as well. So get your brain in order first before you tackle anything else I would say. </s>

<s>[INST] and when something jumps out, I mean that's one of the things that people often ask about is like, well I think there's a bunch of cool stuff out there. Like how do I pick the thing? Like uh how do you, how do you hook in your life? How did you just hook yourself in and stuck with it? [/INST] Well, I got lucky, right? I mean I think that many people forget that a huge amount of it or most of it is luck, right? So um let's not forget that. Um So for me, I fell in love with computers early on because I'm, they spoke to me I guess </s><s>[INST] uh language did they speak [/INST] basic, </s><s>[INST] Basic? [/INST] Um but the uh but then it was just kind of following a set of logical progressions but also um deciding that something that was hard was worth doing and a lot of fun. Right? And so I think that that is also something that's true for many other demands, which is if you find something you love doing, that's also hard if you invest yourself in it and add value to the world, then it will mean something generally right? And again that can be a research paper that can be a software system that can be a new robot that can be that there's many things that that that can be, but a lot of it is like real value comes from doing things that are hard and it doesn't mean you have to suffer, but it's hard. I </s><s>[INST] mean you don't often hear that message we talked about last time a little bit, but it's one of my favorite. Not enough people talk about this. Uh it's uh it's beautiful to hear a successful person well [/INST] and self doubt and imposter syndrome and these are all things that uh successful people suffer with as well, particularly when they put themselves in a point of being uncomfortable, which um I like to do now and then just because it puts you in learning mode, like if you want to, if you want to grow as a person, put yourself in a room with a bunch of people that know way more about whatever you're talking about than you do and ask dumb questions and guess what smart people love to teach often, not always, but often. And if you listen, if you're prepared to listen, if you're prepared to grow, if you're prepared to make connections, you can do some really interesting things. And I think a lot of progress is made by people who kind of hot between domains now and then because they bring uh they bring a perspective into a field that nobody else has, if people have only been working in that field themselves. </s><s>[INST] We mentioned that the universe is kind of like a compiler, uh you know, the the entirety of it, the whole evolution is kind of a kind of compilation maybe are as human beings are kind of compilers. Um let me ask the old sort of question that I didn't ask you last time, which is uh, what's the meaning of it all? Is there meaning like if you asked to compile or why, what would a compiler say? What's the meaning of life? [/INST] What's the meaning of life? Uh you know, I'm prepared for it not to mean anything here. We are all biological things programmed to survive and and propagator our DNA. Um and maybe the universe is just just a computer and you just go until entropy takes over the world and it takes over the universe and then you're done. Um I don't think that's a very productive way to live your life. If so and so I prefer to bias towards the other way, which is saying the world has the universe has a lot of value and I take uh take happiness out of other people. And a lot of times part of that is having kids, but also the relationships you build with other people. And so the way I try to live my life is like, what can I do that has value? How can I move the world forward? How can I take what I'm good at and like bring it, bring it into the world. And how can I, one of those people that likes to work really hard and be very focused on the things that I do. And so if I'm going to do that, how can it be in a demand that actually will matter? Because a lot of things that we do, we find ourselves in the cycle of like, okay, I'm doing a thing. I'm very familiar with it. I've done it for a long time. I've never done anything else, but I'm not really learning. I'm not really, I'm keeping things going, but there's a, there's a younger generation that, that can do the same thing maybe even better than me, right? Maybe if I actually step out of this and jump into something I'm less comfortable with, it's scary. But on the other hand, um it gives somebody else the new opportunity and also then put you back in learning mode And that can be really interesting. And one of the things I've learned is that when you go through that, that first you're deep into impostor syndrome, but when you start working your way out, you start to realize, hey, well there's actually a method to this and, and now I'm able to add new things because they bring a different perspective and this is one of the, the good things about bringing different kinds of people together, diversity of thought is really important. And um if you can pull together people that are coming at things from different directions, you often get invasion. And I love to see that that aha moment where you're like, what we've like really cracked. This is something that nobody's ever done before. And then if you can do it in the context where it adds value, other people can build on it helps move the world. Then that's what that's what really excites me. </s>

<s>[INST] In fact for a limited time. Now if you signed up for an all access pass for a year, you get to get another all access pass to share with a friend, buy one, get one free when I first heard about master class, I thought it was too good to be true, $480 a year, you get an all access pass to watch courses from to list some of my favorites chris Hatfield on space exploration near the Grasse. Tyson on scientific thinking, communication will write the creator of some city and sims on game design, jane Goodall on conservation, Carlos, Santana on guitar, his song, Europa could be the most beautiful guitar song ever written. Garry Kasparov in chess, Daniel Negreanu and poker and many, many more chris Hatfield explaining how rockets work and the experience of being launched into space alone is worth the money For me. The key is to not be overwhelmed by the abundance of choice. Pick three courses you want to complete, watch each of them all the way through. It's not that long, but it's an experience that will stick with you for a long time, I promise. It's easily worth the money, you can watch it in basically any device. Once again sign up at masterclass dot com slash lex to get a discount and to support this podcast and now here's my conversation with David Silver, [/INST] the </s><s>[INST] first program you ever written and what programming language do you [/INST] remember? I remember very clearly, yeah, my, my parents brought home this BBC model B microcomputer. It was just this fascinating thing to me, I was about seven years old and I couldn't resist just playing around with it. Um so I think first program ever um was writing my name out um in different colors and getting it to loop and repeat that. And there was something magical about that which just led to more and more. </s><s>[INST] How did you think about computers back then? Like the magical aspect of it that you can write a program and there's this thing that you just gave birth to that's able to create sort of visual elements and live in its own. Or did you not think of it in those romantic notions? Was it more like, oh that's cool, I can I can solve some puzzles. [/INST] It was always more than solving puzzles. It was something where there was this limitless possibilities. Once you have a computer in front of you, you can do anything with it. Um I used to play with lego with the same feeling, you can make anything you want out of lego but even more so with a computer, you know, you don't, you're not constrained by the amount of kit you've got. And so I was fascinated by it and started pulling out the, you know, the user guide and the advanced user guide and then learning. So I started in basic and then, you know, later 6502. My father was also became interested in the in this machine and gave up his career to go back to school and study for a Master's degree in in artificial intelligence, funnily enough um essex university when I was when I was seven. So I was exposed to those things at an early age, he showed me how to programming prologue and do things like Querying your family tree. And those are some of my earliest memories of trying to um trying to figure things out on the computer. </s><s>[INST] Those are the early steps in computer science programming. But when did you first fall in love with artificial intelligence or with the ideas the dreams of Ai [/INST] I think it was really when I when I went to study at university, so I was an undergrad at Cambridge and studying computer science and and I really started to question, you know, what what really are the goals, what's the goal, where do we want to go with with computer science? And it seemed to me that the the only step of major significance um to take was to try and recreate something akin to human intelligence. If we could do that, that would be a major leap forward. And that idea and it certainly wasn't the first to have it, but it, you know, nestled within me somewhere and and became like a bug, you know, I really wanted to to crack that problem. </s><s>[INST] So, you thought it was like you had a notion that this is something that human beings can do that, it is possible to create an intelligent machine? [/INST] Well, I mean, unless you believe in something metaphysical then, what are our brains doing? Well at some level, they're information processing systems which are okay able to take whatever information is in there, transform it through some form of program and produce some kind of output which enables that human being to do all the amazing things that they can do in this incredible world. </s><s>[INST] So so then do you remember the first time you've written a program that uh because you also had an interesting games, do you remember the first time you were in the program that beat you in a game then? Or beat you at anything sort of uh achieved super david silver level [/INST] performance. So I used to work in the games industry. So for five years I programmed games for for my first job. So it was an amazing opportunity to get involved in a startup company. Um and so I was involved in building a I at that time. Um and so for sure there was a sense of um building um handcrafted what people used to call a i in the games industry, which I think is not really what we might think of as as a i in its fullest sense, but something which is able to um to uh take actions in a way which which makes things interesting and challenging for the for the for the human player. Um And at that time I was able to build, you know, these handcrafted agents, which in certain limited cases could do things which which were able to um do better than me, but mostly in these kind of twitch like scenarios where where they were able to do things faster or because they had some pattern which was able to exploit repeatedly. I think if we're talking about real ai the first experience for me came after that when I realized that this path I was on wasn't taking me towards, it wasn't, it wasn't dealing with that bug, which I still had inside me to really understand intelligence and try and try and solve it. Everything people were doing in games was short term fixes rather than long term vision. Um And so I went back to study for my PhD which was family enough trying to apply reinforcement learning to the game of go. And I built my first go program using reinforcement learning a system which would by trial and error play against itself. Um and was able to learn um which patterns were actually helpful to predict whether it was going to win or lose the game and then choose the moves that led to the combination of patterns that would mean that you're more likely to win in that system. That system beat me? </s>

<s>[INST] I that's kind of awesome. I've never heard about that, that's that's amazing. I'm gonna start doing that now setting alarms like a little bit off t like yeah, I know, I'm never quite sure why the celebration of [/INST] uniqueness. Yeah, and I am quite the odd snowflake in that sense to um So I would usually then try to force myself because I had that same mentality that if I wasn't up at, you know 6, 13 in the gym by seven, that there's something wrong with me and I quickly abandoned that. But if I look back, if there was a shameful act, then I have around my sleep, I think it would be that for some years until I really started to get more detailed into sleep. Um and now I have no shame in telling people that, you know, I will probably usually wake up around 6 45 naturally, sometimes seven when people are looking at me thinking you're a slob, you're lazy. Um and you know, I don't finish my daily workout until, you know, I'm not working until probably nine o'clock in the morning thinking what are you doing now? I will work late into the Day. You know, if I could, I would work 16 hours, it's my passion, just like yours. Um So I don't feel shame around that, but I have changed my mentality around that. </s><s>[INST] Um It's complicated because I um I'm probably happiest going to bed at, if I'm being honest, like at five a.m. [/INST] That's fine, you're just an extreme evening type. But </s><s>[INST] the problem is, it's not that I'm ashamed for. It is uh I actually kind of enjoyed because I get to sleep through all the nonsense of like the morning and isn't that a beautiful thing? Like people are busy with the emails and I just, I'm as happy as a cow. Yeah, and I wake up after all the drama has been resolved and cows [/INST] are happy and the drama has been resolved. </s><s>[INST] Exactly. So, but, you know, in society you do, especially, I mean, this is what I think about is um you know, when you work a larger team, especially with companies, you are, you know, everybody's awake at the same time. So that that's definitely that's definitely been a struggle to try to figure out, just like you said, um how to balance that, how to fit into society and yet be optimal for your prototype, he said, yeah, you have to [/INST] sleep in synchrony with it and harmony. Because normally what we know is that if you fight biology, you'll normally lose and the way you know you've lost is through disease and sickness. </s><s>[INST] You said you suffered through several bouts of insomnia. Is there aside from embracing your prototype, is there advice you can give how to overcome insomnia from your own experience? [/INST] Right now? The best method that we have is something called cognitive behavioral therapy for insomnia or C. B. T. For short. And um you work with people who don't know what it is. You work with a therapist for maybe six weeks and you can do it online by the way, I recommend probably jumping online. It's just the easiest. Um and it will change your beliefs, your habits, your behaviors and your general stress around this thing called sleep. And it is just as effective as sleeping pills in the short term. But what's great is that unlike sleeping pills, when you stop working with your therapist, those benefits last for years later. Whereas when you stop your sleeping pills, you typically have what's called rebound insomnia where your sleep not only goes back to being as bad as it was before, it's usually even worse for me. I think I found a number of things effective. The first is that I had to really address what was stressful and try to come up with some degree of meaningful rationality around it because I think one of the things that happens, there's something very talking about conscious states to come all the way back to um Gosh, I don't know. I feel like we've only been chatting for like 20 minutes, but you're going to tell me it's been a while, it's been a while, okay. I'm desperately, I feel terribly sorry, but let's come back to consciousness where we, where we started. Uh um there is something very strange about the night that thoughts and anxieties are not the same as they are in the waking day. They are worse. They are bigger. And I at least find that I am far more likely to catastrophes and ruminate at night about things that when I wake up the next day in the broad light of day, I think it's nowhere near that bad man, What were you doing? It's not that bad at all. So, to gain firstly some rational understanding of my emotional state that's causing that insomnia was very helpful. The second thing was to keep regularity just going to bed at the same time, waking up and here's an unconventional piece of sleep advice after a bad night of sleep, do nothing. Don't wake up any later. Don't go to bed any earlier. Don't nap during the day and don't drink any more coffee than you would otherwise, because if you end up sleeping later into the morning, you're then not going to be tired at your normal time at night. So then you're going to get into bed thinking, well I had a terrible night of sleep last night and yes, I slept in this morning to try and compensate but I'm still going to get to bed at my normal time. But now you get into bed and you haven't been awake for as long as you normally would. So you're not as sleepy as you normally would be. And so now you sit there lying in bed and it's another bad night. And the same thing is, you know, if you go to bed any earlier, so don't wake up any later, wake up at the same time, don't go to bed in a earlier because then you're just probably your Krone type, your biological rhythm doesn't want you to be asleep. And you think well, terrible night. I'm going to get into bed at nine PM rather than my standard 10. Just gonna be lying in bed awake for that hour, naps will take our double edged sword. They can have wonderful benefits. And we've done lots of studies on naps for both the brain and the body. But they are a double edged sword in the sense that napping, we'll just take the edge off your sleepiness. It's a little bit like a valve on the pressure cooker when you nap during the day. You can take some of that healthy sleepiness that you've been building up during the day and for some people, not all people, but for some people that can then make it harder for them to fall asleep at night and then stay asleep soundly across the night. So the advice would be if you're struggling with sleep at night, don't nap during the day, but if you are not struggling with sleep and you can not regularly naps are just fine and we can play around with optimal durations depending on what you want. Just try not to nap too late into the day because napping late into the day, it's like snacking before your main meal. It just takes the edge off your sleep hunger as it were. But that would be. Um so that's my unconventional 2nd piece of advice regarding insomnia. The 3rd is meditation. I found meditation to be incredibly powerful. I started reading about meditation as I was researching that aspect of the book um many years ago. And as a hard no scientist, I thought this sounds very whoa. Um this is sort of, we all hold hands and sing, come by our and everything is going to be fine with sleep. I read the data and it was compelling. I couldn't ignore it and I started meditating and that was six years ago and I haven't stopped and I found meditation before bed incredibly powerful. The meditation app companies were perplexed at this at first. They want people to meditate during the day. But when they looked at their usage statistics, they found that they would have people in the morning meditating and then there's a huge number of people using the meditation app in the evening. What they were doing was self medicating their, their insomnia and they finally, rather than railing against it, they started to see it as a cash cow. Um, rightly so. So I found meditation to be helpful having a wind down routine is the other thing that's critical for me, I can't just go from because when my mind is switched on and I think you may be like this too if I get into bed that roller decks of thoughts and information and excitement and anxiety and worry is just whirling away and it's not gonna be a good night for me. So I have to find a wind down routine and that makes sense when you realize what sleep is like sleep is not like a light switch. Um sleep is much more like trying to land a plane. You know, it takes time to descend down onto the terra firma that we call sound sleep at night and we have this for kids, you know, I don't have Children, but you know, a lot of parents will say, you know, we have to have the bedroom, sorry, the bedtime routine, you know, you bathe the kids, you put them in bed, you read them a story, you have to go through this routine, This wind down routine for them and then they fall asleep wonderfully, why do we abandon that as adults? We need that same wind down routine. So that's been the other thing that's been very helpful to me. So don't do anything different if you have a bad night of sleep, keep doing the same thing, manage your anxiety, understand it, rationalize it um than meditation and then finally having some kind of disengagement wind down routine. Those are the four things that have been very helpful to </s>

<s>[INST] by the way you got on those shows with no help from them, correct? [/INST] Yeah, that's not even a question. Uh, the reason they would want you to do a book because they know you could get, the only reason people get book deals nowadays literally is because they know that person can market their own book. That's the only way. And I got a Reuben, I gotta Rogan and they don't have the money for the bus for travel, which is fair, they can do Skype. He told me this in writing and I'm like, okay, </s><s>[INST] and they can financially cover Skype, [/INST] but it's like, hey joe, yeah, we don't have the budget, but you're gonna do Skype. Hello. So there is another friend of mine was on a show on CNBC with Nassim Taleb and they said Naseem wants a copy of the book And they're like, oh yeah, it's like 4:00 on Friday. So we're closed. So, and he's like, he went there, picked it up and walked into two blocks. So there is, it's almost cartoonish and it's not incompetence, it's um its past that it's something almost, you can't really believe that I've had two friends who have been literally rendered suicidal um because this is such a huge opportunity for them and they was like watching their kids get beaten in front of them and I had to talk them off the ledge. So it's people do not appreciate how bad, here's another example, </s><s>[INST] the apathy of bureaucracy, something like [/INST] that. I did this book concierge confidential, there's a typo in the first chapter Dance with. I'm about to T. 00. They didn't fix it for the paperback. Okay. It's just like, wow. Okay. Yeah, great book, by the way. Got it. Got NPR gave one of the books of the year. So that was cool. </s><s>[INST] So why participate in this? Because [/INST] otherwise the new york times is gonna pretend you don't exist. Getting book on some books on some shows might be more difficult, although I think that's collapsing in real time. Um you're not gonna get reviewed necessarily places like P. W. Um or some others. </s><s>[INST] So the new book, you're working on your tail, you're white pill. Do I pill? Yeah. Are you self publishing that? Oh yeah, for sure. And what's the thinking behind that? Just because you already have a huge following and a big platform and uh [/INST] it's six times the cash. If I finished the book in december, I could have it out in february if I finished the book in december with the publisher, it's going to be out in December at the earliest 2021. Why am I giving up 10 months of my life? </s><s>[INST] Well this is a big one. Do you have any leverage? Like do do authors have leverage to say f you can you just say what you just look meaning? Like, I want to release this book in two months. [/INST] Oh no, no. I mean, you'll have a contract and then your age you can fight it, but they don't have they don't have the capacity to rush things through. </s><s>[INST] Yeah, I guess if the Because I've heard like big authors, I don't know, Sam Harris, all those folks talk about like, they've accepted it actually. They've accepted their like, yeah, it takes a long time to I'm [/INST] not accepting it, </s><s>[INST] but you but you're kind of implying that a human being like me should, [/INST] like, I'm saying these are your options. </s><s>[INST] All right. So, I just I just hate it. I hate the waiting because it's incompetence. It's not that it's not necessarily the way if I knew it wasn't, you know, if it was the kind of people that are up at two a.m. At night on a friday and they love what you're doing and they're helping create something special. That's the sense I get with some of the netflix folks for example, uh that work with people. I just I don't know anything about this world, but you get like netflix folks who who helped with the shows. You could tell that they're obsessed with those [/INST] shows. You're not gonna get that publishing if you hand, like I handed the book in? I think it was july I didn't hear anything from my editor until december? </s><s>[INST] Well, can we actually talk about the suffering? Sure. The darkest parts of writing a book. So the let's, let's go to the phone, Michael, malice, Stephen King mode of uh, what are the darkest moments of writing this book? And what is it maybe start the white pill? What's the idea? What's the hope and what are your darkest moments around writing this book? [/INST] So, people are familiar with the red pill or the blue pill. The there from the Matrix. The red pill is the idea that what is presented as fact by the corporate press, entertainment industry is in fact a carefully constructed narrative designed to keep some very unpleasant people in power and everyone else under control. And as a one of my expressions is you take one red pill, not the whole bottle because at a certain point you think everything's lie and then you're kind of no capacity for distinguishing </s><s>[INST] Truth. You're full of good one Liners. Well, [/INST] thank you, I'm full of something that's for sure. And what I saw in this space is a lot of these red pilled people got very um disheartened and cynical and one of my big heroes is Albert camus and he said the worst thing of citizens. Um, and that's, that's something called the black pill, which is the idea that, you know, it's it's all, it's just we're waiting for the end. It's hopeless. And I don't see it that way at all. And I'm like, all right, I have to address this and not just with some kind of cheerleading, everything's going great guys here is why I am positive and not that I'm positive, the good guys are going to win, but I'm positive that good guys can win and that's all you need because if you're God forbid kid is kidnapped And there's a 10% chance that you can save them. You're not going to be like, Wow, I don't like those odds, this is your country, this is your values, this is your family. I don't think it's much more than 10%. And even if you lose, you will take pride in that you did everything in your power to win. So is </s>

<s>[INST] So it's a kind of biological game of texas hold em. You're showing some cards and you're hiding other cards. And that's part of it is bluffing and there's and all that. And then there's probably whole species that would do way too much bluffing. That's probably where peacocks fall. There's a there's a book that I don't remember if I read or if I if I wrote if I read summaries of the book? It's about evolution of beauty and birds. Where is that from? Is that a book? Or does Richard Dawkins talk about it. But basically there's some species start to like over select for beauty. Not over select, they just some reason select for beauty. There is a case to be made actually. Now I'm starting to remember, I think Darwin himself made a case that you can select based on beauty alone. So that beauty, there's a point where B doesn't represent some underlying biological truth. You start to select for for beauty itself. And I think the deep question is there is some is there some evolutionary value to beauty? But it's an interesting kind of thought that this can we deviate completely from the deep biological truth to actually appreciate some kind of the summarization in itself. Let me get back to a member because this is a really interesting idea. Um how do a collection of cells remember anything? How do biological systems remember anything? How is that akin to the kind of memory we think of humans as having within our big cognitive engine? Yeah. [/INST] One of the ways to start thinking about by electricity is to ask ourselves, where did neurons and all these cool tricks that the brain uses to run these amazing problem solving abilities on. And basically an electrical network. Right, where did that come from? They didn't just evolve, you know, appear out of nowhere. It must have evolved from something and what it evolved from was a much more ancient ability of cells to form networks to solve other kinds of problems. For example, to navigate more space to control the body shape. And so all of the components of, of neurons. So, so ion channels, uh, neurotransmitter machinery, electrical synapses, All this stuff is way older than brains. Way older than neurons. In fact older than multicellular charity. And so it was already the even even bacterial biofilms. There's some beautiful work from UCSD on, on brain like dynamics and bacterial biofilms. So, evolution figured out very early on that electrical networks are amazing at having memories that integrating information across distance at different kinds of optimization tasks, you know, image recognition and so on. Long before there were brains can </s><s>[INST] actually step back return to it. What is by electricity? What is biochemistry? What is, what are electrical networks? I think a lot of the biology community focuses on the chemicals as them signaling mechanisms that make the whole thing work. You have, I think to a large degree uniquely, maybe you can correct me on that have focused on the bio electricity, which is using electricity for signaling. There's also probably mechanical, [/INST] like </s><s>[INST] knocking on the door. Uh, so what, what what's the difference and what's an electrical network? [/INST] So I want to make sure and and kind of give credit where credit is due. So, so as far back as 1903 and probably late 18 hundreds already, people were thinking about the importance of electrical phenomena in in life. So I'm for sure not the first person to stress the importance of electricity. People there were there were waves of research in the in the thirties, um in the forties and then again in the kind of seventies, eighties and nineties of sort of the pioneers of by electricity who did some amazing work on all this. I think, I think what what we've done that's new, is to step away from this idea, that and I'll describe what what the by electricity is a step away from the idea that well, here's another piece of physics that you need to keep track of to understand physiology and development and to really start looking at this as saying, no, this is a privileged computational layer that gives you access to the actual cognition of the tissue of basil cognition. So, so merging that that developmental biophysics with ideas and cognition of computation. And so and I think I think that's what we've done, that's new. But people have been talking about by electricity for a really long time. And and and so also I'll define that. So, what happens is that if you have, if you have a single cell cell has a membrane in that membrane are proteins called ion channels and those proteins allow charged molecules potassium sodium chloride to go in and out under certain circumstances. And when there's an imbalance of of those ions there becomes a voltage gradient across the membrane. And so all cells, all living cells try to hold a particular kind of voltage difference across the membrane and they spend a lot of energy to do so when you now now. So, so that's that's that's a single cell. When you have multiple cells, the cells sitting next to each other, they can communicate their voltage state to each other via a number of different ways. But one of them is this thing called a gap junction, which is basically like a little submarine hatch, which is kind of docks. Right? And the ions from one side can flow to the other side and vice versa. So </s><s>[INST] isn't it incredible that this evolved is not wild because that didn't exist [/INST] correct. This had to be, this had to be evolved and </s><s>[INST] it had to be invented. Somebody invented electricity in the in the ocean. One of those to get invented. [/INST] Yeah. </s><s>[INST] So [/INST] I mean, it's it is it is incredible. The guy who discovered gap junctions Werner Lowenstein, I visited him. He was he was really old </s>

<s>[INST] did that define, like, help define who you are as a person, that generosity on you. [/INST] It made me think about in my life when it comes to like any sort of business deals or things like that. I don't want to get the best of it in such a way where I screw the other person. I genuinely don't, I'd much rather you owe me than me. Oh you, so if I hire people, they get paid more than they're supposed to and I'd rather than do that and work towards it rather than feel underpaid because if they're underpaid, they'll likely under deliver. Whereas if they feel overpaid then if I need them do something special, they're not gonna be like, I don't get paid for that. Like, yeah, you do. You really do. So that's certainly like played out in my life where I set it up in such a way where I don't owe, you know, I'm owed, but that's okay because I can handle taking the worst of it in spots. I don't like being the person to, you know, I feel like I'm indebted to others. </s><s>[INST] Yeah, and in some way the karma of that tends to pay dividends in the long term. Somehow, somehow there's a, there's somebody up up there that's keeping track in some kind of way. Uh what advice would you give to young people today in high school and college? How to um, have a career that can be proud of, or maybe how to have a life in general, they can be proud of. [/INST] I would say like </s><s>[INST] your [/INST] twenties is a good opportunity to set yourself up for the rest of your life, right? So while the twenties are a period where you want to have fun and you want to experience youth, it's also a good opportunity to start thinking about, what do you want, what do you want to like, what do you want your life to look like in your thirties and forties, Right? So I feel like it's the best time to really put yourself out there and take risks and try to hit it, you know, whatever, you know, like to work really, really hard to set yourself up because, and I said this at a event that was speaking at, when you're like with poker, when your bankroll is very, very small, it's replenish double, right? You don't need to protect it as much as you do. Once you've got something right? Once you have a brand or you have money, you have something like that. That's, that's when you wanna start protecting you. But in your 20's is an opportunity to just really sort of get, you know, to work really, really hard to set yourself up, you know, for the future. I am concerned a little bit like every time I talk to kids today, I'm like, what do you want to be? They all want to be youtubers or instagram stars, rappers, right? Like, okay, it's like, that's cool, but like there's only so many of those that you know that there can be, so it might be worthwhile having a little bit of a backup plan. I </s><s>[INST] think it's easier to be successful on instagram and social media if you do something else. And I [/INST] would say this to one other thing I would say is don't choose a profession or an idea because you think it'll make you rich, right, pursue something that you actually love because if you love it, you're, you're way more likely to become rich. If you don't, you do something that you don't actually enjoy. Now you're spending a lot of your life unhappy doing something you don't want and you're far and if you're not passionate about it, you're probably not the chance of you being successful are much </s><s>[INST] lower and also becoming rich and I've talked to a lot of rich people hang out with a lot of rich people is not going to be as fulfilling as you imagine if you arrive there by not doing the thing that you love doing ultimately. The thing that you love doing is like that's what makes life worth. [/INST] There's another quote, I can't remember who it was, otherwise, I would quote them, but it says something to the effect of like, um if we believe in the lie that more is always better, then we can never truly arrive because wherever we are more is better, right? I've never understood and I've been around rich people, like, you know, you said, you know, I never got, I can't, I don't get it, like if you have a billion dollars, why do you give a shit about money at all? Like, and there's still, like we made this deal and I'm like, you know, we picked up 300 who cares, like your life is set like there is that bell curve right, where obviously being in poverty, you know, there's obviously a high rate of unhappiness, but there's a certain amount of money where you reach, you know, where you reach a level of happiness and then too much you find the people that are searching for money to fulfill these holes, it starts to go back down again. Well, it </s><s>[INST] the getting more money could become a game, like a sport that's fun to play As long as you directly or indirectly acknowledged that what you love the game of it versus the actual attainment [/INST] And I think that's what it is right for me. I've never cared about money that much. I just never did otherwise. I would have a lot more of it but it's always like it's always been strange to me how people that have that kind of money like or cheapen anyway. You know like they wouldn't donate 5000 to a worthwhile charity because it's like buddy this like when it changes your life Not, I don't even like like small things like taxes like Okay you have $20 billion 33%, 31. I get it, I get the, get the point of it all but like it literally has no effect on your life whatsoever. Your life is unchanged whether it's 30. Yeah </s>

<s>[INST] I mean what you have to ask him, did you think deeply about? Because like, you know, you basically have nobody on your team and and so it almost feels like a marriage where is it, what you selective about? Like, James to somebody to bring into your little circle? [/INST] Well, Jamie is exceptional. </s><s>[INST] He is, he truly is a special. I mean he might have grown, I don't remember how he was in the early days, maybe you could say, but [/INST] he was definitely better at it, but he right away. He's exceptional. He's got very little ego, he's not a guy who needs a lot of attention, he's not a guy who um overestimates uh anything like in terms of like negative or positive, like his like his, his interpretation of whether it's uh good things that happen to the show or bad things that happened the show, He just takes it all like flat, he's chill, he's just cool as fun and he's so smart and he's so good as an audio engineer and a podcast producer, he's the best, </s><s>[INST] but he's basically one of the only people on, on this whole team. So yeah, how do you find, I mean when you let people in, I mean, I'm sure other people wanted to get involved, like why don't you have a co host that you basically kind of, well, [/INST] you well, here's the problem with the co host, like when you and I are talking, when we're talking, I'm tuned into you and I'm waiting to hear what you're saying, and I'm listening and I'm interpreting it and then I'm calculating whether or not I have anything to say, whether to let you keep talking, whether I maybe have a question that lets you expand further or whether I have a disagreement or like there's a dance that's going on. Now when there's another person there chiming into it fucks the dance up. It's like dancing. Like if you're doing a dance with someone, you know, like if you're slow dancing with someone and then the third person is they're stepping on his feet. Sometimes it's fun. Sometimes having a third person is fun, comedy podcast. Sometimes it's fun, um fight competitive structure, debate structures, but even then it gets difficult because people talk over each other and also um I find that without headphones, it's way easier to talk over each other. You make mistakes, you don't you don't hear the same way when you have headphones, I hear what you hear, it's all one sound and the audience here is exactly, or rather I hear exactly what the audience here is. Whether it's a over here, my my voice is louder than yours because you're over there and if I don't have headphones on, it doesn't, it's not altogether </s><s>[INST] at that point. One of the interesting things about your show is uh you don't almost never have done and you just generally don't do remote, like um sorry, not remote calls, but you don't go to another person's location, like [/INST] you have only done a few small handful </s><s>[INST] and and just like with the Sapolsky, he should be, he should do this. But actually we went back and forth on email, I told him he needs to get your his as back in in in this in the studio, he's going to work in the book. [/INST] I was a fan of his long time ago because I became obsessed with toxoplasmosis, you know, And uh I've reached out to him a long time ago before he was willing to do it. Then I caught him in downtown L. A. He was there for something else and I just greedily snatched up an hour of his time. </s><s>[INST] Well he doesn't get, I think some of those folks don't get how much magic can happen in this podcast studio, like bigger than anything they've ever done in terms of their work. Not, I'm not talking about reach, but in terms of the discovery of new ideas, there's something magical about conversation like that. Like somebody as brilliant as him. If he gives himself over to the conversation for multiple hours at a time, that's another place where you've been an inspiration where I like, you know, I'm getting more and more confidence of telling people like any long musk that like, you know, a lot of ceos are like, well he has 30 minutes on his schedule. I'm like, no, three hours. So then they're like, so some say no. And then they come back, there's people have started coming back to like, okay, we're starting to get it. They start to get it and you're a rare beacon of hope in that sense that there's some value in long form. They think that Nobody wants to listen for for more than 30 minutes, that they think like I have nothing to say. But the reality is if you just give yourself over to like the three hours just let it go. Three hours, four hours, whatever it is. There's so much to discover about what you didn't even know. You think, [/INST] yeah, you have to be confident that you could do it. And uh, in the beginning I just did it because that's what I wanted to do and no one was listening. So I've always been a curious person So I've always, I've always been interested in listening to how people think about things and how and talking to people about their mindset and just and expanding on my own ideas, just talking sh it and so we would have these podcasts and they would go on forever. And my my friend ari I never let him die, never let this die down. Let let him forget this. He was always like, you have to edit your podcast. I'm telling you right now you're sucking up. I go why he's like, because people are not gonna listen to it. I go they don't have to Yeah, I know you're listening part of it. He was he was just do it just, I'm telling you, trust me, cut it down to like 45 minutes, that's all you need. And I'm like, no, no, I don't think you're right. I I like listening to long form things. No one has that kind of time. Okay, I'm going to do, I'm just going to keep doing it this way so </s>

<s>[INST] continue being a functional agent. [/INST] And and and similarly, if you give any kind of more ambitious goal to an A. G. I. It's very likely to want to acquire more resources so it can do that better. And it's exactly from those sort of sub goals that we might not have intended that some of the concerns about a gi safety come you give it some goal which seems completely harmless. And then before you realize it, it's also trying to do these other things that you didn't want it to do. And it may be smarter than us. So, so </s><s>[INST] fascinating. And let me pause just because I am uh in a very kind of human centric way. See fear of death is a valuable motivator. Uh So you don't think uh do you think that's an artifact of evolution? So that's the kind of mind space evolution created that were sort of almost obsessed about self preservation, Some kind of genetic so you don't think that's necessary to be afraid of death. So not just a kind of sub goal of self preservation. Just so you can keep doing the thing, but more fundamentally sort of have the finite thing like this ends for you at some point, [/INST] interesting. Do I think it's necessary for what precisely </s><s>[INST] for intelligence, but also for consciousness. So for those for both. Do you think really like a finite death and the fear of it is important. [/INST] So before I can answer before we can agree on whether it's necessary for intelligence or for conscience, we should be clear how we define those two words because a lot of really smart people to find them in very different ways. I was in this on this panel with ai experts and they couldn't they couldn't agree on how to define intelligence even so I define intelligence simply as the ability to accomplish complex goals. I like your broad definition because again, I don't want to be a carbon chauvinist, right? And in that case, no, certainly it doesn't require fear of death. I would say Alpha go Alpha Zero is quite intelligent. I don't think Alpha Zero has any fear of being turned off because it doesn't understand the concept of that even and and similarly, consciousness, I mean, you could certainly imagine very simple kind of experience if, you know, if certain plans have any kind of experience, I don't think they're very afraid of dying and there's nothing they can do about it anyway, much to. So there wasn't much value and but more seriously, I think, uh, if you ask, not just about being conscious, but maybe having uh, with you, we we we we we might call an exciting life for you feel passion and really appreciate the things. Maybe they're somehow, maybe there perhaps it does help having having a backdrop today. It's finite. No, let's let's make the most of us live to the fullest. But if you, if you knew you were going to live forever, if you think you would change your </s><s>[INST] Yeah, I mean, in some perspective, it would be an incredibly boring life living forever. So in the sort of loose, subjective terms that you said of something exciting and something in this that other humans would understand, I think as Yeah, it seems that the finite nous of it is important. [/INST] Well, the good news I have for you then is based on what we understand about cosmology. Everything is in our universe's trial, ultimately probably fine. I don't know all the big crunch </s><s>[INST] or, or big what's the expected, the infinite [/INST] Yeah, we could have a big chill or a big crunch or a big rip or that that's the big snap or death bubbles. All of them are more than a billion years away. So we should, we certainly have vastly more time than our ancestors thought. But they're still, it's still pretty hard to squeeze in an infinite number of compute cycles even though there are some loopholes that just might be possible. But I think, you know, some people like to say that you should live as a, you're about you're gonna die in five years or something that's sort of optimal. Maybe we it's good we should build our civilization asset. It's all finite to be on the safe side. </s><s>[INST] Right. Exactly. So you mentioned in defining intelligence as the ability to solve complex goals. So where would you draw a line? How would you try to define human level intelligence and super human level intelligence? Where this consciousness part of that definition? No, [/INST] consciousness does not come into this definition. So, so I think of intelligence as it's a spectrum, but there are very many different kinds of goals you can have, you can have a goal to be a good chess player, a good go player, a good car driver, a good investor, good poet et cetera. So, intelligence that by its very nature isn't something you can measure, but it's one number overall goodness. No, no. There's some people who are better at this. Some people are better than that. Um, right now we have machines that are much better than us at some very narrow tasks like multiplying large numbers fast, memorizing large databases, uh, playing chess playing go soon driving cars. Um, but there's still no machine that can match a human child in general intelligence. But artificial general intelligence A. G. I the name of your course, of course, that is by its very definition. The quest to build a machine, a machine that can do everything as well as we can up to the old Holy Grail of of ai from from back To its inception in the 60s, if that ever happens, of course, I think it's going to be the biggest transition in the history of life on earth. But it but it doesn't necessarily have to wait the big impact until machines are better than I said nothing. That the really big change, it doesn't come exactly the moment they're better than us at everything. The really big change comes first. There are big changes when they start becoming better at us and doing most of the jobs that we do because that takes away much of the demand for human labor. And then the really whopping change comes when they become better than us at A. I research. Right? Right. Because right now the time scale of the AI research is limited by the human research and development cycle of of years. Typically, you know, how long does it take from one release of some software or iphone or whatever to the next. But once, once we once google can replace 40,000 engineers, about 40,000 equivalent pieces of software or whatever. But then and that doesn't, there's no reason that has to be years. It can be in principle much faster. And the time scale of future progress in A I and all of science and technology will be driven by machines, not humans. So it's this point, simple point, which it gives right this incredibly fun controversy about whether it can be intelligence explosion, so called singularity is revenge called it. The idea is articulated by ij good obviously way back 50s. But you can see Alan Turing and others thought about it even earlier. Uh huh. You asked me what exactly what I define </s>

<s>[INST] Apple and [/INST] other such companies in the laptop and desktop space. So the </s><s>[INST] nature of what a computer [/INST] is seems to be changing </s><s>[INST] its like smartphones [/INST] like merging, it's all being a smartphone just with the biggest screen </s><s>[INST] or something like this. So what [/INST] does the future of that look </s><s>[INST] like before [/INST] I answer that question, let me just step back a little bit because and I'm sure we can we can talk more about those things. But the </s><s>[INST] the reality [/INST] Is qualcomm is changing a lot and we use, I know we spent a lot of time talking about 5G and smartphones, snapdragon and I think there has been what had defined quote come for many years but the reality is even consistent with that 5G conversation which is a technology to connect everything. Qualcomm is also changing our technology that was in many cases designed </s><s>[INST] for phones. [/INST] And we said at the beginning connectivity and processing is going to virtually every industry. And as a result </s><s>[INST] Qualcomm is [/INST] really changing with it and expanding to a number of different addressable markets. Some of those markets </s><s>[INST] is [/INST] it's the pc as you talk about it, the the conversions of mobile NPC and the reason I'm excited about this because you see a lot </s><s>[INST] of things happening [/INST] that bring this right front and center when you think about the future </s><s>[INST] technology? So what we learn with the pandemic is [/INST] That the number one </s><s>[INST] use case [/INST] of personal computers </s><s>[INST] is [/INST] communications? </s><s>[INST] Uh It is interesting [/INST] when you think about that, that's the number </s><s>[INST] one use [/INST] case on a pc. Today is communications. </s><s>[INST] It's actually funny [/INST] because in the cellular industry actually, I'll say let me step back in the telecom industry, we've </s><s>[INST] been chasing this killer application of [/INST] video telephony for decades. Right. I remember uh back then in </s><s>[INST] the wire line uh [/INST] even before the internet and I P I S D N. You remember those uh a T and T. Desk phones of a little screen and they said you can do video telephony, we don't watch that. </s><s>[INST] Uh [/INST] in back to the future too. Then when we started developing </s><s>[INST] twiggy said, people said what's the [/INST] application for having dated herself all video telephony. Then </s><s>[INST] We started doing four [/INST] G. And in the beginning people </s><s>[INST] said, well why do you need [/INST] all this broadband? All video telephony. But it took a pandemic to make video telephony the </s><s>[INST] killer application. [/INST] And that's now the number one use case on a pc. So now think about that for a </s><s>[INST] second be [/INST] personal computers. Now there are technologies that people when they were going to buy a pc, They didn't care much about it. Now they do camera, camera? How good is the camera, the audio is that connected? How good is the connectivity do you have the latest and greatest wifi and </s><s>[INST] cellular, what's [/INST] the battery life? Because you're going to be working from anywhere sometimes you know that sometimes you're not. So all those things, what's the </s><s>[INST] portability like? So those things started [/INST] to change how we should think about the pc, but I won't stop there. Let me talk about another trend. So and and all come as a result of what we saw the pandemic, let's say that you are, you're an engineer in the computer aided design. You you have an advance uh </s><s>[INST] desktop [/INST] computer or </s><s>[INST] workstation [/INST] in your office </s><s>[INST] but you [/INST] want to work from home someday, so you're not gonna move that to your home. So </s><s>[INST] what do you need to do? You're going to [/INST] have to rely on that. </s><s>[INST] You're going to run [/INST] that on the cloud </s><s>[INST] and you're gonna run out of the [/INST] cloud that you need high bandwidth Because you almost wanted the cloud to be uh the same as your computer for that use case, that's the 5G on demand computing. </s><s>[INST] Use case [/INST] The use 5G is almost a link between two computers. But then </s><s>[INST] you know, see [/INST] IOS are saying, well my workforce is </s><s>[INST] going home for certain [/INST] days, I want all the data to be in the cloud you look at, for example, Microsoft </s><s>[INST] one drive [/INST] or the ability to collaborate? </s><s>[INST] You need [/INST] the bandwidth. So when you put all of those things together, you start thinking about what is the next generation </s><s>[INST] pc [/INST] and that's the opportunity for qualcomm. I'm just give an example back in uh mobile </s><s>[INST] congress recently [/INST] Lenovo, they have a line of uh, of enterprise laptops called the thinkpad, I'm sure you're familiar with it. So they announced the thinkpad based on snapdragon With 5G on 20 </s><s>[INST] eight hours [/INST] of battery life. So, so that's next generation. </s><s>[INST] It's just a nice screen [/INST] with extremely high </s><s>[INST] nice screen and keyboard uh [/INST] and </s><s>[INST] extremely high connectivity to [/INST] maybe an even more like a more powerful machine </s><s>[INST] in the cloud, something [/INST] more the data connected to the data, continuing to </s><s>[INST] compute all that kind of stuff. [/INST] You have the camera capabilities. And let me </s><s>[INST] go once that more [/INST] Microsoft, talking about some of the </s><s>[INST] features they're doing now [/INST] using </s><s>[INST] on [/INST] Windows 11 Using Snapdragon. </s><s>[INST] Remember [/INST] we talked about a snapdragon has an Ai processor inside </s><s>[INST] there. So one of the [/INST] cool features Microsoft's talking </s><s>[INST] about it is you [/INST] can be on a teams call and you can make sure your eyes are looking at the </s><s>[INST] camera uh percent [/INST] of the time. Well, that's an interesting, </s><s>[INST] so they can be talking [/INST] about that with ai Yes, that's </s><s>[INST] really tricky to [/INST] pull off. For example, the reason I'm a huge stickler for doing these in person, these conversations </s><s>[INST] in person. It's really [/INST] tough to get right, </s><s>[INST] but it's a [/INST] worthy challenge. So that's where the </s><s>[INST] metaverse hopes to. [/INST] So like I just </s><s>[INST] because you said the [/INST] importance of this telephony </s><s>[INST] of humans [/INST] connecting, teleporting </s><s>[INST] themselves. Getting [/INST] that right is really </s><s>[INST] difficult because a lot of people hate zoom meetings, [/INST] but that </s><s>[INST] doesn't mean you can't [/INST] improve that experience </s>

<s>[INST] of conversation with somebody like him. So that story is a story of evolving family dynamics. [/INST] Part of the story of the cartel is individual humans. Marrying other families, getting named the padrino, basically godfathers to other people's kids, forming family and blood ties and influence ties to people not only in Mexico, but in the United States. It's seeing how that dynamic and family dynamic is still there, you know? So he's gone, he's in prison, but he is, he's probably on his way to be our next clandestine saint. You go to the chapel of Malberde. Malberde is basically a Mexican Robin Hood's folk saint down there, who is a saint of traffickers. And at his shrine, you have a small little chapel shrine right next to it. So he's on his way to sainthood in Mexico. You know, not recognized by the Catholic Church, but that doesn't matter in Mexico anymore. Speaking to somebody like him who, you can consider him somebody that lost, you know? He's arrested, but his family's okay. His legacy is out there, he's gonna be named, he's probably gonna be the next folk saint </s><s>[INST] when he passes away, you know? I think he feels like the new wave of what the cartel has become has betrayed him, has left him behind. Or, because it seems like the way the cartel operated [/INST] has changed over the decades. Yeah, well, number one, their power and influence is bigger, you know, they, they are seeing a lower cartel operations in Colombia straight to the, like in the source of it. And then there are clear, they have a clear presence in places like Chicago and Los Angeles. They're in the United States, the whole thought process that a lot of Americans have, like, oh, we don't want that trouble over here, we don't want them to get here, </s><s>[INST] like build a wall and all this, you know? [/INST] So they're deeply integrated into legitimate businesses? I mean, they've been having kids and families up here since for a long time. Some of these people have American passport that works not only directly for them, but have blood ties down there, you know? There's been drag nets and arrests of some of these criminal organizations, states. New generation cartel had one, two, three years ago where, I think it was Operation Anaconda, I think it was called, there was over 80 of their operatives. And this is a new cartel that is very militaristic and growing in Mexico. And they had over 80 arrests in the United States, </s><s>[INST] you know, that, of members of them operating here. And so you could be a legitimate operator inside the United States that's hard to detect. Makes you wonder how many in the US government, the politicians here, is the role of the United States in the drug war, financially in terms of power, is very big. Yeah. [/INST] Surely there's politicians that have a finger into this. Immigration is part of it. Illegal immigration is part of it. And the influence of that has as a bargaining chip and a political chip. We saw this with the first caravan kind of coming up and how it was politicized. The money, fast and furious, and guns being basically let walk down into Mexico. People don't know, basically the ATF had this operation where they were looking at straw purchasers of firearms. Basically people buying up a specific type of firearms that were on a shopping list that the cartels wanted to buy, including, you know, 50 cals, FN-57 pistols, which are small pistols with a high velocity round that will go through a bulletproof vest. AR-15s of all kinds, that could quickly be modified into full auto down in Mexico with drilling a few holes and making a few things to them. So these people were buying all these, the ATF was watching them and allowing them to walk those firearms into Mexico under the guise of trying to track them somehow, which doesn't make a lot of sense for most people to kind of look at that operation. The only people found, the only reason people found out about it was because of the murder of a few federal agents, of the US federal agents that were killed with those guns. One of my friends was shot with one of those pistols outside of his house. And they shot him and they shot his wife. Both of them were killed. Daughter was in the backseat, lost part of her arm. When that happened, the guns were unique. They were like, oh, we didn't ever, the matapolesias is what they call them, they're the cop killers. I hadn't seen those before. So they were unique and interesting. And later on in life, I was watching CNN and seeing the hearings going on. I was like, oh, that's where they came from. Two federal agents changed a lot and it was politicized. There was a whole scandal up here, but in Mexico, how many people died with those firearms being let down, being let go down there? And also what type of sentiment do you think the local populace has of the United States after all those guns were basically handed over to some of these groups? Gun trafficking is another giant part of the equation and part of the problem down there, as far as the amount of munitions, weapons. And now we were also getting trade craft material from conflict zones outside of Mexico. So weaponized drones. The first time we saw some of those weaponized drones was in Syria. And like a few weeks later, grenades were being dropped on the roofs of some public officials building. Cartels are using drones? Yeah, that's been going on for a while. There's a place in Michoacan and has some pretty interesting videos. And the interesting part of them is because the federal police down there are actually working hand in hand with a United Carteles Unidos group, which is basically the local cartels to try and fight off the new generation cartel moving into Michoacan. So even the federal forces are fighting with the cartels to try and keep this larger cartel out. And there's videos of these civilian drones basically dropping explosives. They found some explosive testing ranges out there that are basically replicating stuff that you would see the IRA use during the troubles out there from homemade mortars. IEDs have been used in Mexico, not that much, but they're making like a presence again. We don't have a lot of ordinance around like Iraq, but we do have a big mining industry down there. So mining explosives of all kinds are pretty easy to get. So you start seeing that. And also, I mean, there's some exotic weaponry coming in from the South now and from the ocean. Some of it is probably US military equipment sold to various South American governments that are now not as stable as they were and they're kind of making their way into black market. So a lot of those 50-cal and vehicle-mounted technical-type machine guns and some of the RPGs and man pads or remote-controlled guided missiles that have been found in cartel hands </s>

<s>[INST] Like the push ups and pull ups challenge up and ramping up for It has four flavors, Cocoa, Fruity, Frosted and Blueberry. I tried all of them. They're all delicious. But if you know what's good for you, you'll go with cocoa, my favorite flavor and the flavor of champions. Click the magic spoon dot com slash let's link in the description and use code lex at check out for free shipping and to let them know I sent you, the more you buy for this episode, the more they will support this podcast in the future. So by all of their cereal, it's delicious and good for you. You won't regret it. And now here's my conversation with my dad, alexander Friedmann. What to you is the most beautiful idea in physics. Chemistry or mathematics. Ooh, let's go to that place first. Before we go to the to Kiev [/INST] the most beautiful in physics chemistry. I would say I would say Einstein his ideas regarding statistic motion and brown in motion. I think it is the most beautiful, clear, simple idea because from this like statistics, you can prove existence of molecules. Believe me or not, It was I think it was his first publication in 195 with his first wife actually, it was significantly higher </s><s>[INST] idea. That's right, they were collaborators because those 5445 papers he published [/INST] Yeah, I actually almost everything which Einstein did, it was published the most important stuff He published in one year 1, But the first idea was probably the most elegant. </s><s>[INST] Can you describe brownie in motion and why you think it's beautiful? [/INST] You see that people at that time they understand that if molecules, if they exist, their motion is absolutely chaotic, boom, boom, boom, boom boom, what he proved that doesn't matter what size of particles is their motion is similar and about the same. So it is not necessary to see molecules, you can look at bigger particles, like if you smoke there's particles and you can see like a propagation of this smoke and from here determined behavior of molecules and that's actually relatively short, very elegant, very clear. And at that point people they kept talking about molecules but it was no proof because there is no like a super electronic microscope to see them. So you know, I'm saying yes, molecules exist. You said well, I mean, can you show me there? So I stand made it and boom. And he proved and a lot of people after this publication said probably molecules really exist Einstein 1905. </s><s>[INST] So what do you like about that idea? The simplicity, simplicity, [/INST] that's proving something very very very very, very very complicated. So if I ask you for example, you know, molecules exist can improve it. Obviously, like now you would say, ok, I have a special electronic microscope and I can take a look that small, okay, But when you go back to the beginning of 20th century people, they were laughing, can you prove something? No, you cannot. No, no, no. Take a look at something. Very simple, very, very simple. Very, very simple. </s><s>[INST] What do you think it takes to do that kind of thinking and come to those ideas? And how many times have you encountered those kinds of ideas in your life? Have you had any? Yeah, [/INST] yeah, I would say yeah. And that's uh that's the whole beauty my feeling. That's the whole beauty of science. It's just, you know, people they say no, no, no, no, it's it's impossible. Or and you said no, no, we take a look, take a look, wait a second, read it, take a look. Just think for five minutes if something is very complicated, you know, like a Pizza Nobel prize laureate. Like I worked with him by the way he used to say if it is complicated, it's probably wrong. Well, it was a lot of discussions in capital in Landau because Landau, he he enjoyed to do something in a most complicated way. Why his friend Kapisa, he was trying to do everything is an absolutely simple way. And they had all these discussions because sometimes capital. He went to extremes and he used to say, well this is simple, it means this is right now. He was like, okay, I mean that everything was simple is right. I mean </s><s>[INST] like Feinman to yeah, he liked the simple [/INST] yes, yes, yes, yes. And that's the whole point to find this simplicity. That's that's the beauty of science. That's a beauty really. There's a beautiful and that's something worse to do. Like uh </s><s>[INST] at the risk of going religious for a second. Why do you think our universe allows for such simplicity to be discovered that such simplicity exists at all? [/INST] I can I can tell you it is very similar to Darwin's theory. Why? I don't know why animals are this way? Not the other way. So it is actually a competition. Take a look about simplicity. Uh beauty of physics, like you have to interacting planets and the force is proportional to one over R squared. In this case they are able to go in circles or an ellipsis if there is another law, they are either falling or going out. So like a uh so if like our solar system exists. So just interactions would be one over R squared. That's it. That's very interesting about Isaac Newton and his way of thinking that also was super. I would say genius. </s><s>[INST] Is it weird to you talking about Newton that action at a distance for example, [/INST] this stuff I was doing in elementary school, which is funny without any knowledge of physics, but I was doing physics because of fighting throwing things. Oh, better than that. Better than that we played? Well it was third, 2nd 3rd Grade, So It's in 1960 - 1963. So from one hand, you know Kiev at that time. So people, they were excited about wars obviously all this especially boys and also you know, we were better than americans because you know, we have first sputnik. Yeah and gagarin, I remember how happy we were. When was gagarin? What year? Oh, I'm afraid to make a mistake. But I think it's 1961 or 1960-1961, because what I remember that I wasn't either in the 1st grade or second grade and it was no bread, believe me or not at that time was a big problem because of Khrushchev </s>

<s>[INST] on top of the code or the people in very nice places. Well I think I speak for millions of people yarn and saying thank you for creating this language that so many systems are built on top of them that make a better world. So thank you and thank you for talking today. [/INST] I really appreciate it will make it even better </s><s>[INST] Good. [/INST] </s>

<s>[INST] and specifies. So there's objective function and constraints and variables that becomes separate from the data that operates on. Right? So the that kind of separation allows you to uh you know, put on different hats when put the hat of an optimization person and then put another hat of a data person and dance back and forth. And and also separate the actual uh solvers the optimization systems that do the solving that you can have other people come to the table and then build their solvers whether it's linear or nonlinear uh convex, non convex, that kind of stuff. So of what is the do you as maybe you can comment how you got into that world and what is the beautiful or interesting idea to you from the world of optimization? Sure. [/INST] So I preface it by saying I'm absolutely not an expert on this. And most of the important work in ample comes from why two partners in crime on that bob forer who was a professor of and in the industrial Engineering and Management science department at Northwestern and my colleague at Bell Labs, Dave Gay, who was numerical analyst uh and optimization person. So the deal is linear programming preface this by saying, let's stay with linear program, linear program is the simplest example of this. So linear program is it's taught in school, is that you have a big matrix which is always called A. And you say A X. Is less than or equal to B. So B as a set of constraints. Access the decision variables and a as to how the decision very polls are combined to set up the various constraints. So A as a matrix and X. And be your vectors. Um and then there's an objective function which is just the sum of a bunch of excess and some coefficients on them. And yet that's the thing you want to optimize. Um the problem is that in the real world that that matrix A is a very, very, very intricate, very large and very sparse matrix where the various components of the model are distributed among the coefficients in a way that is totally un obvious to anybody. Um And so what you need is some way to express the original model, which you and I would write, you know, we'd write mathematics on the board, the some of this is greater than the sum of that kind of thing. So you need a language to write those kinds of constraints. And bob for for a long time had been interested in modeling languages, languages that made it possible to do this, there was a modeling language around called Gams, the general algebraic modeling system, but it looks very much like FORTRAN was kind of clunky. Um and so bob spent a sabbatical year at Bell Labs in 1984 and uh he and wasn't in the office across from me and uh it's always geography and uh he and Dave Gay and I started talking about this kind of thing. And um he wanted to design a language that would make it so that you could take these algebraic specifications, you know, summation, signs over sets and that you would write on the board and convert them into basically this a matrix and then pass that off to a solver, which is an entirely separate thing. And so we talked about the design of the language. I don't remember any of the details of this now, but it's kind of an obvious thing, You're just writing out mathematical expressions in a FORTRAN like, or sorry, an algebraic but textual like language. Um and um I wrote the first version of this ample program, uh my first C plus plus program and uh </s><s>[INST] it's written in C plus plus. Yeah. [/INST] And so I did that fairly quickly. We wrote it was, you know, 3000 lines or something, so it wasn't very big, but it sort of showed the feasibility of it that you could actually do something that was easy for people to specify models and converted into something that the solver could work with at the same time as you say, that model and the data are separate things. So one model would then work with all kinds of different data in the same way that lots of programs do the same thing, but with different </s><s>[INST] data. So one of the really nice things is the specification of the models. Human um just kind of like as you say, it's human readable, like I literally remember and stuff, I worked, I would send it to colleagues that I'm pretty sure, never programmed in their life, just just to uh understand what optimization problem is, I think how hard is it to convert that you said you there's a first prototype in people as well as to convert that into something that could actually be used by the solver. [/INST] It's not too bad because most of the solvers have some mechanism that lets them import a model in a form. It might be as simple as the matrix itself in just some representation or if you're doing things that are not linear programming, then there may be some mechanism that you provide things like functions to be called or other constraints on the model. So uh, so all ample does is to generate that kind of thing and then solver deals with all the hard work and then when the solver comes back with numbers, ample converts those back into your original form. So you know how much of each thing you should be buying or making or shipping or whatever. Um, so we did that in 84 and I haven't had a lot to do with it since except that we wrote a couple of versions of the book on which is one of the </s><s>[INST] greatest books ever written. I love that book. I don't know why [/INST] it's an excellent book bob for wrote most of it. And so it's really, really well done. He must spend dynamite teacher </s>

<s>[INST] won't get into architecture like you really, you really want to become physical reality. [/INST] Yes. And I've got close Rome wasn't built in a day. </s><s>[INST] Yeah, but Rome had to start somewhere, I gotta call you out on the advice, there really is a lot of people that look up to you and you've lived a really difficult life and you made it through all of it. What what what do you advise them to do, [/INST] do everything that you feel, even if you are told that it's the wrong thing. No, no, no, no, no, but not anything violent to </s><s>[INST] do what you feel, say what you feel like. Trust your heart, [/INST] Yeah. Trust your gut, trust your instincts. Because that instinct is who God made you to be not. You know what lines up with your like frat brothers or what lines up with what your dad wanted or what lines up with your classmates or you know, it's just God has something inside of you. And when you walk down the street, you're thinking about it and then you just see it in the window and you go down another street, you see the same thing or you're thinking about a number and that number pops up or you see it on a license plate and that's just you know, something is trying to say be the maximum, you know, be the maximum you and God won't give you anything that you can't handle. So for me, everything is a Yes, I think they say be like water. I know Bruce lee said that or matter of fact, just be because we are beings with engineering opportunities </s><s>[INST] beautifully put. Yeah. Uh, I have hope, I have faith that however, this turns out that you have the skill and the capacity to really add love to the world and I hope you do that. I really, really hope you do that. And we talked about the tension of that, the challenges and the opportunities of that. I [/INST] mean I see how we could be of use to each other. You actually have a particular thing for me as I eject from media because even also, you know, I'm really big and like, you know, burning the phoenix to all, you know, as an engineer that yourselves completely regenerate every seven years. I think I, I rushed the process a little bit so I run around looking like a burn victim sometimes, </s><s>[INST] but I gotta tell you anyone I, I'm close with, I work with, there has to be trust, there has to be love and I think you've been burned quite a bit in your life [/INST] sometimes. I did it myself, I put my hand on the stuff so how do we get to engineering? There's a deal to be done here, not saying you're making this money off of that and making that money off, but there's same twins. So there's dealings here, you got pain, I got pain but were beings with engineering opportunities </s><s>[INST] for a long time. I thought somebody should create a new social network and it looks like a great engineering could do that. You [/INST] got to at one time from two of your homeboys at one time. </s><s>[INST] Yeah. And that's a great way. I mean every, you know, it can be done from scratch or it can be done a new with an already established platform. Yeah, I'm very excited by that possibility parlors smaller, twitter is larger. I wonder whether you can do any kind of revolutionary engineering with a large company. If anyone can, I would say it's Ellen, [/INST] well, Ellen can look at what we do over here. That's why I like God does this a lot of times. Uh, he creates doppelgangers. There's a lot of examples where theory happen two places at one time. A really similar type of personality happens at one time. It's </s><s>[INST] weird. It's beautiful, hilarious actually. It's kind of humorous. [/INST] So we should, we should just share every piece of information together as like as I go bill parlor and as he builds up twitter and see how to make both those better. But I'll be like, hey, I'm not gonna let Snapchat bully me out of, you know, adding something to it. That's best. It should just be the best platform possible. That means that, you know, that means Tiktok has amazing features, twitter has amazing things about it. I know this girl that's like, I like instagram better than twitter is just she likes that twitter has porn but other than that instagram but I was like, well instagram is pouring also it's like soft, it's more subtle. It's not telling you it's not, it's not conservative by any means, you know, and I </s><s>[INST] mean that's the thing, you have to please a lot of people with these platforms and I think the goal of a platform should be to help individuals maximize their long term happiness and growth, like intellectually spiritually, psychologically, all of that. It's a tricky thing to figure out how to do that. Of course you could just have a platform just, just strictly for fun. I think that's where Tiktok leans, but um, there has, I think there's a balance to strike like twitter. There's a lot of intellectual philosophical stuff going on on twitter, but it's somehow currently makes it too easy to be divisive and hateful towards each other and cancel and all that kind of stuff. [/INST] I think we need to have these healthy conversations. That's what it is. Healthy conversation healthy dialogues. That's the biggest thing. Like when I first wore the red hat, the conversation is less healthy than it is today. The conversation about George Floyd. You know, it's like, did I send the tweet the right way. Did is Candace Owens documentary. Perfect. Is what Michael moore is that documentary is perfect century self isn't fully accurate. You know, it's 2-3 hours of a transmitter of information through uh, particularly usually one person's view and anything you want to research if you're interested in it, You gotta look at 10 versions of it. 20 versions of it, George Floyd, we've looked at one version of that now, we have a second. Well, by the time we get to the you know, I know this is like, it's it's not my last hoody but it's definitely not my first hoodie that I put on. You know, it's like, that's how we got to like even this particular size of zipper or the fact that, you </s>

<s>[INST] So you're also, you're widely acknowledges, I mean, outside of your science work as being one of the greatest educators in the world. And Feinman is famous [/INST] Yeah, </s><s>[INST] for for being that, is there something about being a teacher you've well, [/INST] it's it's very, very rewarding when you have students who are really into it. And you know, going back to Feinman at caltech, I was taking these graduate courses and there were two of us, myself and Jeff Richmond who's now a professor of physics at University of California santa Barbara who asked lots of questions and a lot of the caltech students are nervous about asking questions. They want to save face. They seem to think that if they ask a question their peers might think it's a stupid question. Well, I didn't really care what people thought and Jeff Richman didn't either. And we ask all these questions and in fact in many cases they were quite good questions and Feinman said, well the rest of you should be having questions like this. And I remember one time in particular when he said, you know, he said to the rest of the class, why is it always these two aren't you aren't the rest of you curious about what I'm saying? Do you really understand it all that well? If so why aren't you asking the next most logical question? I know you guys are too scared to ask these questions that these two are asking. So he actually invited us to lunch a couple of times, were just the three of us sat And had lunch with one of the greatest thinkers of 20th century physics. And so yeah, he, he rubbed off on me and you </s><s>[INST] encourage questions [/INST] as we invert courage questions, you know? And uh, yeah, definitely, I mean, you know, I encourage questions. I I like it when students ask questions, I tell them that they shouldn't feel shy about asking a question, probably half the students in the class would have that same question if they even understood the material enough to ask that question. </s><s>[INST] Yeah, curiosity is the first step of uh, absolutely. Of seeing the beauty of something. So yeah, the, and the question is the ultimate form of curiosity. Yeah, let me ask, what is the meaning of life? [/INST] The meaning of life, you know, </s><s>[INST] from a cosmologist perspective or from a human perspective, from my [/INST] personal, you know, life is what you make of it really right? It's each of us has to have our own meaning and it doesn't have to be, well, I think that in many cases, meaning is to some degree associated with goals, you set some goals or expectations for yourself, things you want to accomplish things. You want to do things you want to experience and to the degree that you experience those and do those things. It can give you meaning. You don't have to change the world the way Newton or Michelangelo or da Vinci did. I mean people often say change the world, but look, come on, there's 7.5 close to eight billion of us now, most of us are not going to change the world. And does that mean that most of us are leading meaningful lives? No, it just has to be something that gives you meaning that gives you satisfaction. That gives you a good feeling about what you did and often based on human nature, which can be very good and also very bad. But often it's the things that help others that give us meaning and a feeling of satisfaction. You taught someone to read, you cared for someone who was terminally ill. You brought up a nice family, you brought up your kids, um, you did a good job, you put your heart and soul into it. You read a lot of books, if that's what you wanted to do, had a lot of perspectives on life. You, you traveled the world if that's what you wanted to do. Um, but if some of these things are not within reach, your in a socio economic position where you can't travel the world or whatever you find other forms of meaning. It doesn't, it doesn't have to be some profound. I'm going to change the world. I'm going to be the one who everyone remembers </s><s>[INST] type [/INST] thing right? </s><s>[INST] In the context of the greatest story ever told. Like the fact that we came from stars well and now or to apes asking about the meaning of life, how does that fit together? Well, does that make any [/INST] sense? You know that it does, It does. And this is sort of what I was referring to that it's a beautiful universe that allows us to come into creation, right? It's a way that the universe found of knowing of understanding itself because I don't think that, you know, inanimate rocks and stars and black holes and things have any real capability of of abstract thoughts and of learning about the rest of the universe or even their origins. I mean, they're just, they're just a pile of atoms that's that's has no conscience, has no ability to think, has no ability to explore. And we do. And you know, I'm not saying we're the epitome of all life forever, but at least for life on earth so far, the evidence suggests that we are the epitome in terms of the richness of our thoughts, the degree to which we can explore the universe. Do experiments, build machines, understand our origins and I just hope that we use science for good, not evil and that we don't end up destroying ourselves. I mean the whales and dolphins are plenty intelligent. They don't know yet ask abstract questions. They don't read books. But on the other hand, there not in any danger of destroying themselves and everything else as well. And so maybe maybe that's a better form of intelligence, but at least in terms of our ability to explore and make use of our minds. I mean to me it's this it's this that gives me um the potential for meaning. Right? The fact that I can understand and explore. </s>

<s>[INST] The following is a conversation with No Brown research scientists at fair facebook ai research group at meta ai. He co [/INST] created the first </s><s>[INST] Ai system that achieved superhuman level performance in no limit texas. Hold'em both heads up and multiplayer. And now recently he co created an Ai system that can strategically out negotiate humans using natural language in a popular board game called diplomacy, which is a war game that emphasizes negotiation. And now a quick few second mention of each sponsor. Check them out in the description is the best way to support this podcast. We got true classic tees for shirts audible for audiobooks. Inside tracker for bio monitoring and express VPN for I S. B privacy choose wisely, my friends and now on to the full ad reads as always no ads in the middle. I try to make this interesting, but if you skip them, please still check out the sponsors. I enjoy their stuff. Maybe you will too. This show is brought to you by true classic tees. High quality, soft, slim fitted t shirts for men, they also make all the other menswear staples, like polos, workout shirts and boxers. I mostly wear their black t shirt and that's what I'm wearing right now, I have a ton of them and you know how it feels and how it looks. That's like the standard default look of the guy behind the keyboard. I guess programmer, I wonder when the t shirt was created. Looking up on the wiki the t shirt evolved from undergarments used in the 19th century and in the mid 20th century, transition from undergarments to general use casual clothing. Okay, wonderful. Anyway, go to true classic dot com and enter code lex to get 25% off. This episode is brought to you by audible audiobook service. That has given me hundreds if not thousands of hours of education to listening to audio books. I'm currently re listening for the, I don't know how many times I've re listened to this book, how many times I've read it, but George Orwell's 1984 not nearly as many times as Animal Farm. I don't know why exactly, but Animal Farm just as a story, just connects with me a basic fable of creatures being not so good to each other and being greedy and also being good to each other and having personalities and just in a simple way that sort of uh whenever animals are made in the Children like story, it's simple, but it can still get to the profound and I think I like to think of life in those terms. Anyway, I'm currently listening to 1984. It's really well, sort of spoken voice or whatever on audible. So I highly recommend it. New members can try free for 30 days at audible dot com slash lex or text lex to 505 100. This show is also brought to you by inside tracker. A service. I used to track biological data. Did you know? There's 1.2 to 1.5 gallons of blood in a human being. It's approximately 10% of adults way. I remember like reading some insane stats on the volume of blood pumped a day through the human body. It's kind of, it's just incredible, sort of this whole machinery that interacts with oxygen. I mean, just with the chemistry of Earth interacts with it. It uses energy ultimately from the sun and the uh, photosynthesis is involved with the plans. The whole like system is just freaking incredible and it's just so full of data. Like imagine being able to measure all of that about life on Earth. Yeah. So inside trackers, like early steps and being able to measure directly from the human body, get $200 off Inside Trackers, Ultimate plan or 34% off the entire store. When you go to Inside Tracker dot com slash lex. This show is also brought to you by express VPN. I use them to protect my privacy on the internet. I've used them for many, many years through thick and thin when I was lost in this two big world of ours and looking for the truth out there, protecting my privacy all along. I remember being really excited to learn that it can work on android, that it can work on Linux. So it works on any operating system and it just, it doesn't feel like it's there. Everything is super fast whether you have the VPN or not. It's kind of incredible. It's incredible how this technology works and it conceals your location from uh the perspective of the website or service that you're using online. It's it's an incredible tool. Should be the first layer of protection of your privacy on the internet. Go to express VPN dot com slash lex pod for an extra three months free. This is a lex Friedman podcast to support it, please check out our sponsors in the description and now Dear friends, here's no [/INST] Brown, </s><s>[INST] you've been a lead on three amazing ai projects. So we've got laboratories that solved or at least achieved human level performance on no limit texas hold'em poker with two players. Heads up, [/INST] you got </s><s>[INST] pluribus that solved? No limit texas hold'em poker with six players. And just now you have Cicero, these are all names of systems that solved or achieved human level performance on the game of diplomacy which for people who don't know is a popular strategy board game. It was loved by JFK, john f, Kennedy and Henry Kissinger and many other big famous people in the decades since. So let's talk about poker and diplomacy today. First poker, what is the game of no limit texas hold'em and how is it different from chess? [/INST] Well, no limit texas hold'em poker is the most popular variant of poker in the world. So you know, you go to a casino, you play, sit down at the poker table, the game that you're playing is no limit texas hold em if you watch movies about poker, like Casino Royale or rounders, the game that they're playing is no limit texas hold'em poker now, it's very different from limit hold'em in that you can bet any amount of chips that you want and so the stakes escalate really quickly, You start out with like one or $2 in the pot and then by the end of the hand you've got like $1000 in there, maybe to </s>

<s>[INST] That's beautiful. That's beautiful. That that because the change you will have in the world that you don't know for sure, but it'll happen across, uh years, even decades. So you have to think in the future, like you're working in the future? Your actions are working in the future and a lot of people will judge this conversation and others you've done and so on in the president and that doesn't matter. [/INST] You know why I apologize When I say, I'm sorry to the jewish people that I hurt. It's not about, like I said, oh, I said, I'm not gonna hold this apology hostage, right? It's obviously it's about God, but God is everything, right? So what's the point of having language and being, you can't engineer. If you said God God God, right? It's like, it's God asks us to be judgmental. So that you become, if you're a doctor, right? It's your judgment that saves grandma, right? And God puts that in this, like we have God inside us, we're a piece of God also, right? So I'm gonna go specific to a person that story about da Vinci alone is enough of a reason for me to give a sincere apology to the jewish people because I do believe that there's a da Vinci code inside of all of my misspellings and scribbles right here, that God wants me to get to the world. And when I'm in my way, if God has set something on me, when I'm in my way, I'm in God's way, now I can point at other people and say, hey, you guys are in God's way because you're not listen to me. But if I'm in my way, I'm not listening to God and for me to be a philosopher and a leader or whatever other language, I have to listen to God. So before God, what I would do is start off as a samurai and say, I'm sorry for hurting you as a jewish person. I'm sorry for the way that made you feel and I'm sorry for the entire uh population of a race that I feel is actually my brothers because I classify and feel that I'm also connected with christ in that way that my people came from africa and that way and I can't say it's this exact teaching or that exact teaching, but I feel that there is an important, you know, the sons of Abraham, however we want to word it right for us to come together and bring our different talents together to serve God collectively. And he as much of an alien as I am, he does not cause me, he did not call for me to alienate and that's what the problem has been, especially in politics and in America, he plays America as the tv as the radio to the world. We invented rock and roll, right? We are the influencers of the rest of the planet. We're the, we're the youngest startup in history, right? We're the youngest country, we're the, we're the youngest superpower, right? So for me in my position, the fact that I can talk to trump one day and be at a game with kim Kardashian the next day and then be at paris fashion week. They don't want me to be involved with. Language of decisive divisive language, you know, proving that I can be just as racist as a racist white person. Oh, I can, I can be a racist black person also. That's not what God has called me to do. So how do I simplify and say, hey to you as a visionary, as a, as an engineer, as a person. That's given me a platform that, as my brother, I want to apologize to you for how that made you feel and to the rest of the people that carry specifically that pain, not the pain that I talk about, That doesn't seem to be as much of a pain for people when I talk about the abortion clinics and stuff, that's it's like that's the argued pain. Even in my own community, like not everyone thinks that's a pain, but for the jewish community, it's collectively this was a pain. And I've gone, I've gone so far, just psychologically on this journey since that statement had happened where I didn't want to like give up the candy in my pocket. I didn't want to say, hey, I stuck my hand in the candy jar. I wanted to defend my freedom of speech. I wanted this. But even with freedom of speech, pain is pain causing people to hurt is not helping, causing people to hurt is not helping. So that means I have to take a more sophisticated approach to engineering this problem. This problem is the opportunity. And yes, that what could some people may deem as a mistake? Would it be more genuine if I called it a mistake? I think I would be hypocritical by calling a mistake with the ending of that. The idea I'm gonna say out loud. So I'm not, I don't try to be a hypocrite on purpose. Uh, so, but I will call it a mistake to add a genuine, uh, to be genuine about the feeling, with that mistake, with me as a human being, not to justify, but I have to state that God makes no mistakes because if I hadn't made that mistake, I wouldn't be in front of you today and that's not to justify this </s><s>[INST] uh [/INST] sleepy text. But </s><s>[INST] I can tell you're a man that has love in his heart. And I hear that I hear it through the words explicitly and implicitly. And I think if we're to engineer a better future, the way to do that is with love. So as one human to another, I love you brother. Thank you for talking today. This is great. [/INST] I like you </s>

<s>[INST] No, the the chest [/INST] upbringing. I make it come close to you, right? You see like whenever I do my exercises, the vector is always pulling straight towards me. Yeah. So just uh, you know, cupping close to you, the most dangerous thing that a person can do to me on an arm wrestling match is just pull me away from my body. That's that's a terrible thing for me. Yeah. So so that cupping, that's a massive part of the sport. So now when you think, what does the cup due to the other person, if I cup, they get turned over, right? So this has to get really strong this pro nation. So </s><s>[INST] so to fight the to fight that rolling. [/INST] Exactly, yeah. So that's that's through the thumb. </s><s>[INST] Yeah. Oh, so you put uh you got it, you put on the thumb and you put this motion. Yeah. Got it. [/INST] And those two things, those two things together, this cupping and rolling. This is what's going to make the person's hand bend back and once a person's hand is bent back there, just their whole game gets cut to pieces. They have very little good options. It's all like nasty stuff. Yeah. So those two things that's a that's a huge part of of your of your investment, uh rise always be climbing. Yeah, exactly. Yeah. Those three simple things, that's what I would tell anybody to spend most of their time on if you want to become an arm wrestler. </s><s>[INST] So, and to use bands would be good for this. [/INST] Bands are great because they're easy to transport. The only problem I have with bands is like, if you like to measure, you know, if you like to be precise, bands just aren't that precise </s><s>[INST] to have growth. Yeah, [/INST] just, I mean, it's just like, you know, you know exactly what you need. The prescription is kind of a band is kind of like and a lot of people myself included. I like to um I like to know exactly my outputs. So so weights, </s><s>[INST] so it would, it would be like cables, [/INST] cables are nice. Bands are great too. I I mixed the two bands are uh when I kind of don't need to they're more like easy for me. When I trained bands, bands are dangerous because the acceleration is so high on them. Like when you screw up with band training, the acceleration is way faster than gravity. Right? So if you do something bad, I can, I can make it go really much worse. </s><s>[INST] It's funny that you didn't mention bicep curls or it's a chain. It's changed. So you're I mean the idea if you focus on these three, the other stuff catches up, like it's all involved, this whole this whole thing is involved. [/INST] If you have an ax, right? The blade of the ax, that's that's these things, right? Like you need the pointy end of all your attacks to be awesome. </s><s>[INST] If you have a super sharp ax. Yeah, you can have a shitty hands. Yeah. Right. Yeah. So focus on that. The the point, the tip of the acts, yeah, the tip of the ax [/INST] is so important, right? Like if I have an awesome bicep and I can't quite use it. What's it good </s><s>[INST] for? Right? Yeah. I think a lot of emotions with the risk that you mentioned are just thinking about jujitsu, especially in the G. Um there's a lot of, I mean, there's so much importance to this and people don't often work it explicitly. So many of the chokes require ability to, it's almost like exactly like arm wrestling very close because you're weak here. What's that called, floppy wrist and you're strong with the cup. Yeah. And so you just getting the most of whatever that's involved, the muscle, the turning the pressure because that's where also the choke comes that little, the thing that makes you win an arm wrestling is also the thing that finishes the person when you have them grabbed the strength is very similar. Yeah, it's fascinating actually, of course, like you said, if you want to be very good, you should be doing the very specific exact emotion. [/INST] So if I was going to do jujitsu, I'd be like working out with the g </s><s>[INST] Yeah, the problem is, you know, it's difficult to construct the, the exact, so you have to actually go with people and then they don't like being choked right? So like, it's hard to, I'm actually a big, we have these kinds of debates all the time. Is um you know, I'm a big believer in drilling, I love doing something thousands of times like john donna hurt somebody. And I mentioned to you about the judi stick folks here, they're less believers in drilling. They, they see the value of almost like the mind of going live and exploring ideas. It's that play. You don't need to do the thing 1000 times. You just need to always be thinking about the little details that make you make you better and then in action practicing, like developing the strength, the power of the explosion of the agility in action. So actually rolling, I don't, you know, I, I agree with this, but I just believe in volume more so [/INST] you can accomplish it through volume. You can, you can play a lot. </s>

<s>[INST] You know, I was actually, somebody brought this up to me. Uh, I mean, all forms of light are even, uh, even like the ones that are like big, you can make look, I mean, you make look beautiful like that. There's no sense in which you can't integrate it into design. Like there's all kinds of awesome designs. I don't think small and humble is beautiful. It could be like, you know, brutalism or like it could be uh, like harsh corner. I mean, like I said, like hot rods. Like, I don't like, I don't necessarily like, like, man, I'm going to start so much controversy with this, but I don't like Portia's okay. The Porsche 9 11, like everyone says it's the most beautiful. No, no, it's like, it's like a baby car. It doesn't make any sense. But everyone, it's beauty's in the eye of the beholder. You're already looking at me like what this kid talking [/INST] about, your happy to talk about, you're digging your own hole. The form and function and my take on the beauty of the hardware that we put on our vehicles. You know, I will not comment on the course monologue. </s><s>[INST] Okay. All right, so, but aesthetics fine. But there's an underlying like philosophical question behind the kind of lighter question is like how much of the problem can be solved with computer vision, with machine learning. So I think without sort of disagreements and so on. It's nice to put uh, it on the spectrum because when I was doing a lot of machine learning as well, It's interesting to think how much of driving, if we look at five years, 10 years, 50 years down the road, what can be learned in almost more and more and more and to end way if you look at what Tesla is doing with as a machine learning problem. They're doing a multi task learning thing where it's just they break up driving into a bunch of learning tasks and they have one single neural network and they're just collecting huge amounts of data that's training that everything hung out with George shots. I don't know if, you know George. Uh, I [/INST] love him so much. </s><s>[INST] He's just an entertaining human being. We were off mic talking about Hunter s Thompson. He's the hunter? S Thompson of autonomous driving. Okay, so he I didn't realize this with comma ai but they're like really trying to end to end, they're the machine the machine, like looking at the machine learning problem there. Uh really not doing multitask learning, but it's uh it's computing the drivable area as a machine learning task. And hoping that like down the line, this Level two system that's driver assistance will eventually lead to allow you to have a fully autonomous vehicle. Okay. There's an underlying deep philosophical question. Their technical question of how much of driving can be learned. So lighter is an effective to today for actually deploying a successful service in phoenix, right, That's safe, that's reliable etcetera etcetera. But uh the question and I'm not saying you can't do machine learning and lighter, but the question is that like how much of driving can be learned eventually? Can we do fully autonomous that's learned. [/INST] Uh you know, learning is all over the place and plays a key role in every part of our system. I as you said, I would decouple the sensing modalities from the ml and the software parts of it. Uh lighter radar cameras. It's all machine learning, all of the object detection classification of course, like that. That's what, you know, these modern deep nuts incontinence. Very, very good at you feed them raw data, massive amounts of raw data. Um and you know, that's actually what our custom build lighters and Raiders are really good at radars. They don't just give you point estimates of objects in space. They give you raw my physical observations and then you take all of that raw information, there's colors of the pixels whether it's, you know, lighters returns and some auxiliary information. It's not just distance, right? And you know angle and distance is much richer information that you get from those returns plus really rich information from the Raiders. You fuse it all together and you feed it into those massive animal models that then uh lead to the best results in terms of you know, object deduction classification state estimation. </s><s>[INST] So there's a side to interrupt but there is a fusion. I mean that's something that people didn't do for a very long time which is like at the sensor fusion level I guess like early on fusing the information together whether so that the sensory information that the vehicle receives from the different modalities or even from different cameras is combined before it is fed into the machine learning models. [/INST] Uh Yes, I think this is one of the trends you're seeing more of that. You mentioned intend there's different interpretation of intend, there is kind of the purest interpretation that I'm going to like have one model that goes from raw sensor data to like you know, steering torque and you know, gas breaks, you know that that's too much. I don't think that's the right way to do it. There's more, you know, smaller versions of intend where you're kind of doing more entry in learning or core training or d propagation of kind of signals back and forth across the different stages of your system there's been a really good ways it gets into some fairly complex design choices where on one hand, you want modularity and the company compressibility compressibility of your system. Um but on the other hand, you don't want to create interfaces that are too narrow or too brittle to engineered, where you're giving up on the generality of the solution or you're unable to properly propagate signal, you know, rich signal forward and losses and you know, back back so you can optimize the whole system joins me. Uh So I would decouple and I guess what you're seeing in terms of the fusion of the sensing data from different modalities as well as kind of fusion in the temporal level going more from, you know, frame by frame where, you know, you could have one net that would do frame by frame protection on camera and then, you know, something that does frame by frame and lighter and then radar and then you fuse it that, you know, in a weaker engineered way later the field over the last decade has been evolving in more kind of joint fusion more end to end models that are solving some of these tasks, you know, jointly and there's tremendous power in that. And you know, that that's that's that's that's the progression that you kind of our, you know, our stack has been on as well Now it's your you know that so I would decouple the kind of sensing and how that information is used from the role of the male and the entire stack and you know, I guess it's uh, there's trade offs, uh, and you know, modularity and how do you inject inductive bias into your system? Right. This is uh, there's tremendous power in being able to do that. So we have, there's no part of our system that is not heavily that does not heavily, you know, leverage data driven development or state of the art ml but it's mapping, there's a simulator or there's perception, you know, object level, you know, perception, whether it's semantic, understanding prediction, decision making, you know, so and so on. Um, it's and of course object detection and classification, like you're finding pedestrians and cars and cyclists and you know, cones and signs and vegetation and being very good at estimating kind of detection classification of state estimation. There's just table stakes like, like that's step zero of this whole stack. You can be incredibly good at that whether you use cameras and light as a radar, but they're just, you know, that's table stakes. Just stop zero beyond that. You get into the really interesting challenges of semantic understanding the perception level, you get into seen level of reasoning, you get into very deep problems that have to do with prediction and joint prediction and interaction social interaction between all the actors in the environment, pedestrians, cyclists, other cars and get into decision making, right? So how do you build a law systems. So, uh we leverage ml very heavily in all of these components, I do believe that the best results you achieved by kind of using a hybrid approach and having different types of ml uh having uh different models with different degrees of inductive bias that you can have uh and combining kind of model free approaches with some model based approaches and some uh rule based physics based systems. So, one example I can give you is traffic lights. Uh There's a problem of the detection of traffic lights state and obviously, that's a great problem for, you know, computer vision confidence are, you know, that's their bread and butter, right? That's how you build that. But then the interpretation of, you know, of a traffic light that you need to learn that, right? You read you don't need to build some complex ml model that infers with some, you know, precision and recall that red means stop. It was it's a very clear engineered signal with very clear semantics. All right. So, you want to induce that bias, like how you induce that bias and that whether, you know, it's a constraint or a cost function in your stack, but like, it is important to be able to inject that, like clear semantic signal into your stack and, you know, that's what we do. Um And but then the question of like and that's when you apply it to yourself, when you're making decisions whether you want to stop for a red light, you know or not, uh but if you think about how other people treat traffic lights were back to the ml version of that, you know, they're supposed to stop for a red light, but that doesn't mean they will. So then you're back in the like, very heavy uh ml domain where you're picking up on like very subtle keys about, you know, they have to do with the behavior of objects and pedestrian, cyclist, cars and the whole thing, you know, entire configuration of the scene that allow you to make accurate predictions on whether they will in fact stop or run a red light. </s>

<s>[INST] America. And when everybody, not everybody unmasked, there's a large number of people who have become softer and softer capitalism creates an environment where there's people that still step up amidst that with a big dream and challenge the conventions that human spirit just to rise above that is the last example that Jeff Bezos is an example that and and they push, you know, the limits of those of human beings that are willing to step up and you know, I I you know, I think about sort of how to create a company that uh, that amidst all of the softness still creates a revolution. It's not, it doesn't seem trivial. It seems like how do you build a culture that's once healthy but also unhealthy [/INST] in the world. It's all </s><s>[INST] time pursuit is, [/INST] it's all top down everybody just, you asked earlier what leadership wasn't never answered the question I think you know, what leaders do is they set the example, they set the bar And if you look at someone like Elon, you know, we're personal friends for 20 years. Um, and he is indefatigable. Like, I mean the guy has a stamina that is just phenomenal, like he does not get tired. He works relentlessly and he sets that standard for the rest of the team. And I think you know Bezos is very sharp and like to debate stuff and it's very, you know and jobs was just incredible at design and figuring out how to bridge that gap. So they just, leaders set the standard, they set the standard and you know, that your time is over as a leader when you can't set the standard and that's when you have to pass the baton, right? And besides did that. And besides now is saying, you know what, I'm 57 I'm the richest guy on the planet, Uh, depending on the week and uh, I would like to do some other challenges, but I don't want to grind it out at Amazon for another 25 years. I want to do other things. And so he passed the baton and that's the healthy thing to do in that regard. I do think there is a time period in which you can run that hot and then at a certain point you have to then change, just like an athlete might go to be a coach, right? And you're a commentator. And so, you know, being an entrepreneur is brutal. It's, you know, seven days a week, 12 hours a day, anybody who says anything differently is kidding themselves. You're going to have to sacrifice if you in this competition in America has to fight. If America does not win capitalism and china does, it is literally the end of the human species. It's over for humanity right now, everything has been going really well in terms of the number of people living in poverty is plummeting life. Uh, you know, lifespans have been rising, science is booming, The economy is booming. All these things are incredible. The one thing that's kind of stagnant right now is the number of people living in democracy versus Under authoritarian rule. It's flat. So when you look at all Steve Pinker charts and he's really excited, there's one you're going to see this flat and I think we peaked with 53 or 54% of people on the planet Earth being in a democracy and now it's going below 50 and it's because some of the democratic, you know, Western countries don't have the population growth of some of the communist and socialist countries, an authoritarian countries. And we have to make sure that we're we win capitalism, we must win economically. That is the battlefield, the battlefield of science, technology and money and economy. Finance, that's the battlefield china wins, authoritarians win. And at any time, Jinping can pull Jack Ma into a room and say it's time for you to be reeducated or they can put three or four million people leaguers into prison camps and say, you know what, this religious thing that's counter to what is productive for us. Therefore we're gonna shave your heads and we're going to have you literally pick cotton in the fields. They have, we Gers with no sense of any kind of arc of history in the fields picking cottons as slaves in what can only be described by every humanitarian organization as a concentration camp and every jewish person I know takes great offense when somebody uses the holocaust as a metaphor, except in the case of the weavers right now. And every jewish person I've talked to has said to me that is a holocaust that is millions of people going to genocide because of their religious beliefs and I'm an atheist. But if people want to believe a certain religion fine, but you know, china's approaches, we need to win capitalism so bad. We need to win on the global stage. So bad. We can't have any of this religious stuff going on here. That is a distraction from winning and beating America. And then in America, the people who are going to make us win or the entrepreneurs and the scientists and the technology and our education system and finance and we're vilifying those things. It's pretty dark, </s><s>[INST] it's dark. But I still believe that the vilification is just in the space of twitter in the space of idea. I think that's probably a good entrepreneurs went out in the end. They don't believe in that and they'll bill, we'll get the [/INST] rocks. Some of them do actually in their darkest moments. I can tell you that they turned off their twitter accounts and they, I've had to sit down with a number of entrepreneurs and say turn off twitter, this is not healthy for you, this is not a healthy pursuit because I don't read the comments if you do, it's like a full contact sport, you should just take it as like professional wrestling or something, but stay focused on building companies. Uh, and you know, advancing the human species through science and technology </s>

<s>[INST] years. Do you think those mistake? Yes. What do you think that was the motivation behind saying that other companies are also promising L5 vehicles With their different approaches in 2020, 2021, 2022. [/INST] If anybody would like to bet me that those things do not pan out. I will, I will bet you even money even money. I'll bet you as much as you want. So </s><s>[INST] are you worried about what's going to happen because you're not in full agreement on that? What's going to happen when 2020-21 come around and nobody has fleets of autonomous vehicles? [/INST] Well, you can look at the history if you go back five years ago, they were all promised by 2018 and 2017 </s><s>[INST] but they weren't that strong of promises. I mean ford really declared pretty I think not many have declared as as like definitively as they have now these dates? [/INST] Well, okay, so let's separate L four and L five. Do I think that it's possible for Waymo to continue to uh Kind of like like hack on their system until it gets to level four in Chandler Arizona. Yes. </s><s>[INST] When there's no safety driver [/INST] Chandler Arizona </s><s>[INST] by what? By which year are we talking about. [/INST] Oh, I even think that's possible by like, 2020 2021 But level four Chandler Arizona not level five New York City </s><s>[INST] level for meaning some very defined streets. It works out really well, [/INST] very defined streets and then practically these streets are pretty empty if most of the streets are covered in way most Waymo can kind of change the definition of what driving is right? If your self driving network is the majority of cars in an area, they only need to be safe with respect to each other and all the humans will need to learn to adapt to them now go drive in downtown new york. Yeah that's already, you can talk about autonomy and like like on farms already works great because you can really just follow the GPS line. </s><s>[INST] Yeah. So what does uh success look like for comma ai what what are the milestones like where you can sit back with some champagne and say we did it boys and girls. [/INST] Um Well it's never over. </s><s>[INST] Yeah, but it was drink Champagne every show what is good. What are some winds? [/INST] Um A big milestone that we're hoping for uh by mid next year is profitability of the company. Um and we're going to have to revisit the idea of selling a consumer product but it's not going to be like the common one when we do it, it's going to be perfect. Open pilot has gotten so much better in the last two years we're gonna have a few few features, we're gonna have 100% driver monitoring. We're gonna disable no safety features in the car. Um actually I think the really cool what we're doing right now, our project this week is we're analyzing the data set and looking for all the http triggers from the manufacturer systems. We have better data set on that than the manufacturers. How how much does, how many does Toyota have 10 million miles of real world driving to know how many times they're a be triggered. </s><s>[INST] So let me give you because you asked right uh financial advice because I work with a lot of automakers and one possible source of money are you, which I'll be excited to see you take on is basically selling the data. So which is something that most people and not selling in a way we're here here automaker but creating, we've done this actually at M. I. T. Not for money purposes, but you can do it for significant money purposes and make the world a better place by creating a consortium to where automakers would pay in and then they get to have free access to the data. And I think a lot of people are really hungry for that and we pay a significant amount of money for it. Here's [/INST] the problem with that. I like this idea. All in theory would be very easy for me to give them access to my servers and we already have all open source tools to access this data. It's in a great format. We have a great pipeline, but they're gonna put me in the room with some business development guy and I'm going to have to talk to this guy and he's not going to know most of the words I'm saying. I'm not going to tolerate that. </s><s>[INST] Okay. Mick Jagger. No, no, no, no. I think I agree with you. I'm the same way. But you just tell them the terms and there's no discussion needed. If I [/INST] could, if I could just tell them the terms and then like all right, who wants access to my data, I will sell it to you for uh let's say you wanna, you wanna subscription. I'll sell you for 100 came up Any 100 k month, 100 a month. I'll give you access to the data subscription. Yeah, I think that's kind of fair came up with that number off the top of my head. If somebody sends me like a three line email where it's like we would like to pay 100 a month to get access to your data. We would agree to like reasonable privacy terms of of of of the people who are in the data set. I would be happy to do it. But that's not going to be the email. The email is gonna be, hey, do you have some time in the next month where we can sit down and we can, I don't have time for that. We're moving too fast. You </s>

<s>[INST] Is it possible to return to a place where there is friendly competition between nations that ultimately unites and inspires the peoples of these different worlds, these very different worlds that have especially recently come to conflict over the over the war in Ukraine, the tension builds, the war, the conflict, the suffering is actually creating more and more division, creating more and more hate. I think as we've talked about, I think science and engineering, and especially the most epic version of engineering, which is rocketry and space travel, unites people, unites people even in a time of tension, conflict and war. So do you have a hope [/INST] that we can return to that place? I think historically, spaceflight has been one of the most bonding things. You know, we look at we have countless examples of, you know, Cold War enemies coming together and working together, lending a hand. Apollo 13, for example, of course, you know, there is the potential that who knew where is going to reenter since it was not in the plane trajectory at all for reentry. And the Soviet Union said, hey, wherever, you know, wherever they land, like, we'll help you guys out, basically, you know, that was a pretty big thing at the time, obviously. We also in 1975 saw the Apollo Soyuz mission, which was an Apollo spacecraft docking with the Soyuz spacecraft. First time there's international collaboration. And again, 1975 still very admits the Cold War. Yet we have this collaboration that I don't know what else could have done that, you know, I mean, and think about what actually takes to do that you have to come up with a docking module that is like that takes the two different air environments and the two different docking systems and talk to the engineers and mission planners and figure out, you know, train together the cosmonauts and the astronauts trained together and got to know each other, they were crossing boundaries and borders, and coming together for this mission. And even if it was totally a fluff piece, like, Like even if it was totally this like, you know, cynical, you know, just trying to make a pretty face for everybody, for the cameras or something, </s><s>[INST] obviously it still had an impact. Yeah. It had the symbolic impact, but there's also the practical impact. I mean, a lot of people have to work together. Yes. And that has a ripple effect on the culture, on the different engineers. [/INST] Yes. Yeah. 100%. And even just the astronauts and the cosmonauts involved, like think about what probably went through their heads during this process of like going from, oh my God, I'm going to have to work with them to getting to know them and then sharing meals in space. Like that's a crazy transformation of timelines. And I would love, I do think that spaceflight has the ability to bond us and unite us because it is ultimately, you know, this little tiny little planet we're floating around on, you know, it's the only, it is the boundary that we all share, you know, you only can, it only takes you getting off of this planet to realize, oh my God, we're all neighbors. We're all living in the same house together. And I do think ultimately, you know, as we continue to expand our horizons and expand our exploration, that it has the potential to unite us more </s><s>[INST] than it has the potential to divide us. So one of the potential conflicts of the 21st century that I think everyone wants to avoid, both in the cyberspace and in the hot war space, cold war, hot war, all kinds of war, all kinds of economic conflict this was between the United States and China. So China is going full steam ahead in developing a space program, doing a lot of incredible work. Like you mentioned, 64 launches in 2022 with two failures, but you know, moving straight ahead. [/INST] And by the way, the affiliation, and by the way, the affiliation, we had a lot of startups, like a lot of the launches were from brand new companies. So to have two failures out of 64, I mean, that's still, </s><s>[INST] if you look at operational launches, it was flawless. Do you see a pathway where there's again, in that same way collaboration or a friendly competition between all the different companies and nations in the United States and China in the next, as we push towards the moon, Mars and beyond. [/INST] I held a dumb hope that China would actually be allowed to sign onto the Artemis Accord, to be able to take part in this next step towards the moon. I mean, just imagine if like, if they provided a, you know, a propulsion module or a land or something, and we actually came together to land on the moon, instead of having another space race, you know, it's like, it would have been so cool. And yeah, I still am hopeful that similar to, back in the Cold War, that we might have something like that someday, where we actually are collaborating. And it feels like sometimes we're really close to that and other times it feels like we're really far from that. And it just sucks because, I know, and you know, I try really hard on my channel to always separate and celebrate the work being done, you know, because at the end of the day, there's someone that's just going home to their family, clock in an hour is working really hard on pushing their program and doing engineering work. And it's, we don't get to choose where we're born and what we're born into. So I really like to avoid the political aspects of things and the geopolitical aspects and just appreciate the science and the science we're seeing and the progress that China's doing in the last 10 years is very akin to early spaceflight programs and with the runway of like, just keep on going. Like I see no reason for them to be slowing down. So it's definitely something to watch and be interested in and who knows. I mean, there really genuinely might be a race to the moon again </s>

<s>[INST] percent is always the problem because that's probably why I hesitate to touch conspiracy theories because I'm allergic to certainty in all forms in politics in any kind of discourse. And people are so sure in both directions actually it's kind of hilarious either they're sure that the conspiracy theory particular, whatever the conspiracy theories is false. Like they almost dismiss it like uh like they don't even want to talk about. It's like the people like the way they dismiss that the Earth is flat. Most scientists are like, they don't even want to like hear what the what the flat earthers are saying. They don't have a zero patience for it. Which is like maybe in that case it's deserved. But everything else you really like have empathy. Like consider the you have like a this is weird to say, but I feel like you have to consider that the earth might be flat for like one minute. Like you have to be empathetic, you have to be open minded. [/INST] I don't see a lot of that there are cultural taste makers and more and that's that really is what concerns me the most because it's just another manifestation of all of our problems is that we have this completely bifurcating economy bifurcating culture literally in terms of We have the middle of the country and then we have the coast and in terms of the population, it's almost 5050. And with you know, increasing megacities in urban culture like urban monoculture of L. A. New York and Chicago and D. C. and Boston and Austin relative to how an entire other group of americans who live their lives or even the people within them who aren't rich and upwardly mobile, how they live their lives is just completely separating and all of our language and communication in mass media and more is to the top and then everybody else has forgotten. </s><s>[INST] Do you think when you go, when you dig to the core, there is a big there's a big gap between left and right. Is there is that division that that's perceived currently real or most people look center left and center [/INST] right? It's so interesting because that's such a loaded term center left. What does that mean? Like? So you I think the way you're thinking of it is I'm not like a well even this like I'm not a radical socialist, but I'm uh marginally left on cultural issues and economic issues, this is how we've traditionally understood things. Um and then when when in popular discourse like center right, like what does it mean to be sent to write? Like I am marginally right on social conservative on social issues and marginally right on economic issues. But that's just not possible. Like if you look at survey data for example, like uh stimulus checks, People who are against stimulus checks are conservative, right? Well, 80% of the population is for a stimulus check. So that means a sizable number of republicans are for stimulus checks. Same thing happens on like a wealth tax. Um The same thing happens on, okay, florida voted for trump 3.1 More than Barack Obama 2008, the same day passes a $15 minimum wage at 67%. So what's going on? So that's why what is going on? Well, that's what my career, </s><s>[INST] but it seems like so that's that's fascinating conversation is different than the policies. [/INST] Well, it's different than reality. That's what I would say. Which is that the way we have to understand american politics today. It didn't always used to be this way is it's almost entirely along basic. I would say the main divider is because even when you talk about class, this misses it in terms of socio economics, it's around culture, which is that it's basically if you went to a four year degree granting institution, you are part of one culture if you didn't, you're part of another, I don't want to erase the 20% or whatever. People who did go to college degree who are republicans or vice versa, et cetera. But I'm saying on average, in terms of the median way that you feel we're basically bifurcating along those lines because people get upset be like, oh well, you know, there are rich people who voted for trump. And I'm like, yeah, but you know, you know what they are, they're like plumbers or something. Like there are people who make $100,000 a year, but they didn't go to a four year college degree and they might live who are in a place which is not an urban metro area. And then at the same time you have like a vox writer who makes like 30 grand, but they have a lot more cultural power then like the plumber. So you have to think about where exactly that line is. And I think in general that's the way that we're trending. So that's why when I say like what's going on? Are we divided? Yeah, like but it's not left and right. I mean like, and that's why I hate these labels. So </s><s>[INST] it's more like it's more just red and blue teams, their arbitrary teams. So how arbitrary are these teams? I guess it's another completely arbitrary. So you kind of imply that there's, I don't know if you're sort of in post analyzing the patterns? Because it seems like there's a network effects of like you just pick the team red or [/INST] blue </s><s>[INST] and it might have to do with college, you might have to do all those things. But like it seems like it's more about just the people around [/INST] you correct. </s>

<s>[INST] Well, I mean the question is how much of the game of life can be turned into. [/INST] So that's that I think is a really good question. And I think that we don't I don't think we as a I don't know, community really know that the answer to this, but um so okay, so so I went I went to a talk by some experts on computer chess. So in particular, computer chess is really interesting because, you know, for of course for 1000 years, humans were the best chess playing things on the planet. Um and then computers like the head of the best person and they've been ahead ever since. It's not like people have have overtaken computers, but um but computers and people together have overtaken computers. So at least last time I checked, I don't know what the very latest is, but last time I checked that there were teams of people who could work with computer programs to defeat the best computer programs </s><s>[INST] in the game of go [/INST] in the game of chess, in the game of chess. And so using the information about how these things called elo scores, this sort of notion of how strong a player are. You, there's there's kind of a range of possible scores and the you you increment in score, basically if you can beat another player of that lower score, 62% of the time or something like that. Like there's some threshold of if you can somewhat consistently beat someone, then you are of a higher score than that person. And there's a question as to how many times can you do that in chess? Right. And so we know that there's a range of human mobility levels that cap out with the best playing humans and the computers went a step beyond that computers and people together have not gone. I think a full step beyond that. It feels the estimates that they have is that it's starting to assume toad that we've reached kind of the maximum the best possible chess playing. And so that means that there's kind of a finite strategic depth right? At some point you just can't get any better at this game. </s><s>[INST] Yeah, I mean, I don't uh so I'll actually check that. Uh I think it's interesting because if you have somebody like uh magnus Carlsen who's using these chess programs to train his mind like to learn about chess player. And so like that's a very interesting thing because we're not static creatures. We're learning together. I mean, just like we're talking about social networks, those algorithms are teaching us just like we're teaching those algorithms. So that's the fascinating thing. But I think the best just playing programs are now better than the pairs. Like they have competition between paris, but the, it's still, even if they weren't, it's an interesting question, where is the ceiling? So the David, the ominous David Silver kind of statement is like we have not found the ceiling. [/INST] Right. But so the question is, okay, so I don't I don't know his analysis on that. My from talking to go experts. The depth, the strategic depth of go seems to be substantially greater than that of chess. That there's more kind of steps of improvement that you can make, getting better and better and better better, but there's no reason to think that it's infinite. Uh And so it could be that it's the what David is seeing is a kind of assume toting that you can keep getting better. But with diminishing returns and at some point you hit optimal play. Like in theory all these finite games, they're finite. They have an optimal strategy. There's a strategy that is the mini max optimal strategy. And so at that point you can't get any better. You can't beat that strategy. Now that strategy maybe from an information processing perspective, intractable, right? That you need all the situations are sufficiently different, that you can't compress it at all. It's this giant mess of hard coded rules and we can never achieve that. But that still puts a cap on how many levels of improvement that we can actually make. </s><s>[INST] But the thing about self play is if you if you put it, although I don't like doing that in the broader category of self supervised learning is that it doesn't require too much or any human human Yeah. Human label or just human effort. The human involvement past a certain point and the same thing you could argue is true for the recent breakthroughs in natural language processing with language [/INST] Models. This is how you get to gpt three </s><s>[INST] Yeah, see how they did the uh [/INST] that was a good, good transition. </s><s>[INST] Yeah, I practiced that for days leading up to this guy now. Uh but like that's one of the questions is can we find ways to formulate problems in this world that are important to us humans? Like more important than the game of chess, that to which self supervised kinds of approaches could be applied, whether it's self play, for example, for like maybe you could think of like autonomous vehicles in simulation, that kind of stuff, or just robotics applications and simulation or uh in the self supervised learning, we're a nanite stated data or data that's generated by humans naturally without extra costs, like the Wikipedia or like all of the internet can be used to learn something about to create intelligence systems that do something really powerful that passed the turing test or that do some kind of super human level performance. So what's your intuition trying to stage all of it together about our discussion of g the limits of self play and your thoughts about maybe the limits of neural networks in the context of language models? Is there some intuition in there that might be useful to think about? [/INST] Yeah. Yeah, yeah. So so first of all, the whole transformer network family of things um is really cool. It's really, really cool. I mean, you know, if you've ever back in the day you played with, I don't know, Markov models for generating text and you've seen the kind of text that they spit out and you compare it to what's happening now. It's it's amazing. It's so amazing now. It doesn't take very long interacting with one of these systems before you find the holes, right? It's it's not smart in any kind of general wei it's really good at a bunch of things and it does seem to understand a lot of the statistics of language extremely well and that turns out to be very powerful. You can answer many questions with that, but it doesn't make it a good conversationalist, right? And doesn't make it a good storyteller. It just makes it good at imitating of things that is seen in the past. </s>

<s>[INST] So you're responsible not just for a set of particular habits and so on. You're responsible for everything. [/INST] Yes. And your respond, here's here's the big one, right? You're responsible for shaping your own soul, your consciousness, you get to decide what it's going to be like. </s><s>[INST] And the only tool you have is your mind. [/INST] Your only reason is your mind. Well, your emotions play a tool when they're properly cultivated. They play a role in that. And the tools you have is thinking, experiencing living, coming to the right conclusions. You know, listening to great music and and watching good movies and and and and art is very important in shaping your own soul and helping you do this. It's got a it's got a crucial role in that bet it's work and it's lonely work because this work you do with yourself. Now, if you find somebody who you love, who shares these values and you can do with them, that's great. But it's my only sleet lonely work. It's hard. It's challenging. It ends your world. The reward is unbelievable. But but even at the think about think about the enlightenment, right? So up until the enlightenment, where was truth. Truth came from a book. And there were a few people who understood the book, most of us couldn't read, and they conveyed it to us and they just told us what to do. And in that sense, life's easy. It sucks and we die young and we have nothing and we don't enjoy it, but it's easy. And the enlightenment comes around and says, we've got this tool, it's called reason. And it allows us to discover truth about the world. It's not in the book, It's actually your reason allows you to discover stuff about the world. And I consider the first really the first figure of the enlightenment is Newton not lock right. It's a scientist. Because he teaches us the laws of mechanics. Like how does stuff work? And people go, oh, wow, this is cool. I can use my mind. I can discover truth. Isn't that amazing? And everything opens up once you do that. Hey, if I can discover if I understand the laws of motion, if I can understand truth in the world, how come I can't decide who I marry? I mean, everything was fixed in those days. How come I can't decide what profession I should be in, right? Everybody belonged to a guilt? How come I can't decide who my political leader should be. Mm hmm. That's so it's all reason. It's all. Once you understand the efficacy of your own mind to understand truths, understand reality, discovered you would not understand it. Discovered everything opens up. Now. You can take responsibility for your own life. Because now you have the tools to do it. But we are living in an era where postmodernism tells us there is no truth. There is no reality and our mind is useless anyway. Critical race theory tells us that you're determined by your race and your race shapes everything, and your free will is meaningless and your reason doesn't matter because reason is just shaped by your genes and shaped by your color of your skin. It's the most racist theory of all. And you've got you've got our friends at U. C. Irvine telling them your senses don't tell you anything about reality. Anyway, reality is what it is. You know what's the purpose of reason? It's to invent stuff. It's gonna make stuff up. Then. What use is that? It's complete fantasy. You've basically got every philosophical intellectual voice in the culture telling them their reason is impotent. There's like a steven pinker who tries and I love Pinker and he's really good and I love his books but you know he needs to be stronger about this. And there's a few people are kind of there's a few people partially in the intellectual dark web and otherwise who are big unreason but not consistent enough and not full understanding of what it means or what it implies. And then there's little old </s><s>[INST] me, it's [/INST] me against the world in a sense because I'm not only willing to accept to articulate the case for a reason. But then what that implies, it implies freedom and implies capitalism, it implies taking personal responsibility over your own life and the other intellectual dark web. People get to reason and then oh politics, you can be whatever. No you can't you can't be a socialist and for reason it doesn't actually that those are incompatible. And you can't be a determinist and for a reason. Reason and determinism don't go together. The whole point of reason is that it's an achievement. It requires effort and requires engagement requires choice. So it is it does feel like a little old me because that's that's it. The allies, I have allies, I have allies among the some libertarians over economics have some allies in the intellectual dark web, maybe over reason, but none of them are allies in the full sense. My allies are the other objectives, but we're just not a lot of us. So </s><s>[INST] for people listening to this for the few folks kind of listening to this and and thinking about the trajectory of their own life. I guess the takeaway is the reason is a difficult project, but a project that's worthy of taking on. [/INST] Yeah, difficult is I don't know if difficulties doing because difficult sounds like it's, you know, I have to push this boulder up a hill. It's not difficult in that sense. It's difficult in the sense that it requires energy and focus. It requires effort, but it's immediately rewarding. It's fun to do and its rewards. Uh, immediate uh, pretty quick, right? It takes a while to undo all the garbage that you have. But we all have that I had. It took me years and years and years to get rid of certain concepts and certain emotions that I had. That didn't make any sense. But it takes a long time to fully integrate that. So </s>

<s>[INST] So, you think that there is uh just like joseph Campbell, there's a thread that connects all of these stories, narratives that were constructed for ourselves as we evolve and that threat is grounded in some kind of absolute ideas of maybe on the morality side, which is the trickiest one of good and evil [/INST] somewhat. Yeah, I think that a lot of this stuff is just derived from a biological perspective. I feel like these things are innate within us, </s><s>[INST] I think are innately humans are good. Like we [/INST] know, I don't, I feel like I also feel like there's an issue of scale to like um like Nassim Taleb likes to talk about how he views his the way he interacts with with groups in terms of scale. You know, what is this thing about like at uh at the familiar level, I'm a I'm a communist and then at the the civic level, I'm a republican or something that this other level. And then it goes on at the widest level. He's a libertarian or something of that nature, you know, </s><s>[INST] like fundamentally human interaction changes [/INST] on scale, on scale and scale and also from uh you know, subjective to the environment around them. So, and I don't even mean environment, just in the sake of physical environment, nature, right? Like nature is constantly trying to murder you. Well, it's not really trying, it's just nature's being nature university of the Universe and at times it takes you out. It's just not with any particular uh compunction or prejudice, it just loops, you know, I'm sorry there's no more dodos my bad </s><s>[INST] but don't you think the particular flavor of the complexity that is the human mind was created? Like let me make an argument for that all people are fundamentally good is uh there's an evolutionary advantage to being the striving to cooperate to add more love to the world of like compassion empathy, all that kind of stuff and that the very thing that created the human mind was this evolutionary advantage, whatever the forces behind this evolutionary advantage and [/INST] scale. Yes. So when we're dealing with a small tribe sure when you meet another tribe, maybe there's other factors that are going to into that, let's say you scale up and so your 1 50 has exceeded their 1 50. And like you start to get to a certain point where um you can't really be close enough to someone down the line of some of that next like that 1 51 51 50 and they just now all of a sudden become some some guy whatever. And when it comes to some guy at once it starts hitting scale, I don't know that it's capable. People can be as as magnanimous to a stranger as to the known if they orient themselves to be secure enough because it does come to security insecurity. And one way or the other either brought on by the unknown brought on by an actual threat brought on by even their own as we would use the word insecurity in that, their own insecurity within their own capabilities, Their own belief in themselves, all these things can change things from being compassionate and what have you to at least at the very least maybe not evil but self interest driven to the point of negative results for those that aren't, you know what I mean? </s><s>[INST] But another way to frame that is uh maybe it's less about scale and more about the amount of resources available. So if we're overflowing with resources in terms of ah security and safety, all the uh things you mentioned, if we have more than enough resources than the way we treat a stranger, the way we position ourselves towards that stranger might be uh in a way that allows us to be our real human selves as opposed to sort of our animal self. And therefore it's mostly about how clever can we, descendants of apes be in coming up with all cool kinds of technologies and ways to uh officially use the resources we have such that we're not constrained and my hope is that we can that human innovation will outpace the growth of our the number of people that are starving for resources. [/INST] Yes, I think that there's a lot of uh rationality behind uh this argument and you know in some ways I agree and a lot of ways I see it as missing the point of of how this experiment has been playing out across time. When you look at what what for one it's like define resources, you know, what is a what is a resource of as humans uh would would define it right? Or wealth even. And so you can say well you know an iphone is a resource, the internet is a resource uh water obviously is a resource, but if we weigh them, what is more important to human beings. Water internet or iphones, it's water, right? So if we look at resources, if we start with, what do human beings need to live? I mean actually live not live here in this bullshit fantasy creation extension of our own ingenuity and you know, a prison of our own creation and also a paradise of our own creation. But this is not how human beings normally live. This is all built upon stuff on, it's built on concept on idea and some and some of it's built on just well this is the paradigm. So this is what you do. Human beings need food, they need water to survive. They need shelter from the elements and they need certain skills to perpetuate these things and be able to pass them down so that they can So these things don't become, you don't end up in this this gap where you have to relearn things because if it's lost than that time before you can get it back again is going to be a dark ages of sorts, you know, or it's going to be highly detrimental to your group because not knowing how to fish, not knowing how to hunt, not knowing how to even clean and cook the game once you have, it could be lethal. </s>

<s>[INST] skin and being able to [/INST] laying it out efficiently. </s><s>[INST] Like is there, it sounds absurd, but you [/INST] could roll it up, it would be, </s><s>[INST] I mean, or you could just lay outside [/INST] longer. I wanted to think if you just need to have one hour and how much time would you need? But if you just went out there for </s><s>[INST] four hours you need something [/INST] that's smaller. But you know I think of you know sort of a half </s><s>[INST] a tennis courts so you could [/INST] make like wings could be like and you lay </s><s>[INST] them out there and but but also if [/INST] you that's if you need </s><s>[INST] all your energy [/INST] only from your skin. </s><s>[INST] So if you just get a little bit of [/INST] it and your energy of course you could just walk around with your skin as is and you still have to eat but not as much. And I </s><s>[INST] describe that because [/INST] we need other ways to think about making your own energy. If you're in a really </s><s>[INST] long mission that's far from [/INST] stars, you could turn on a lamp that would give you some of that. You know, </s><s>[INST] essentially exact [/INST] wavelength of light you need for your chloroplast and your skin. </s><s>[INST] But it's you know, that's something [/INST] I'm hoping would happen in 3-4 </s><s>[INST] 100 years. But it would [/INST] be hard because you're taking a </s><s>[INST] plant organelles [/INST] and putting it into animal </s><s>[INST] cells. Which [/INST] sounds weird but we have mitochondria inside of us. </s><s>[INST] But basically we ourselves [/INST] captured bacteria and now it walks around with us all the time. So </s><s>[INST] there's precedent for an evolution. [/INST] How much by the way speaking </s><s>[INST] of which [/INST] does evolution help us here? So </s><s>[INST] we talked a lot about [/INST] engineering, you know, building, </s><s>[INST] you [/INST] know, genetically modifying humans </s><s>[INST] to make them [/INST] more resilient </s><s>[INST] or having [/INST] mechanisms for repairing </s><s>[INST] parts of [/INST] humans. </s><s>[INST] Um What [/INST] about evolving humans or evolving organisms that live on </s><s>[INST] humans? Sort of the [/INST] thing you mentioned, </s><s>[INST] which you've already [/INST] learned is that humans are pretty adaptable. </s><s>[INST] And what does that [/INST] mean? You also um somewhere wrote that, you know, there's trillions of cells that </s><s>[INST] make up the human body and you know, those are [/INST] all organisms and there </s><s>[INST] are also very [/INST] adaptable. So </s><s>[INST] can we leverage the natural process [/INST] of evolution </s><s>[INST] of the [/INST] slaughter, the selection, </s><s>[INST] the mutation and the adaptation [/INST] that is all. And uh sorry </s><s>[INST] to throw slaughter into there. Just just acknowledging [/INST] that a lot of organisms </s><s>[INST] have to die in [/INST] evolution. </s><s>[INST] Uh can [/INST] we use that for long term </s><s>[INST] space [/INST] travel or </s><s>[INST] colonization occupation? [/INST] Is there a good word for this? Of </s><s>[INST] of planets [/INST] like terraform the planet, you know, to adjust the </s><s>[INST] human body to the planet? [/INST] Oh it's not really a term for that yet. I guess too </s><s>[INST] adapt [/INST] to uh the </s><s>[INST] new vacation spot. [/INST] Yeah, I called it </s><s>[INST] just directed [/INST] evolution in the book is that you you you guide the evolution towards what you want in this case. Sometimes you can engineer yourselves to make exactly what you want. But other times you put people on planets and </s><s>[INST] and see how they change. [/INST] Actually later in the book. I imagine if you have humans on multiple planets you could have this virtuous cycle or as people adapt and evolve here you'd sequence their DNA and see how they change and then send the information back to the other planet and then study them with more resources. So you'd be able to then </s><s>[INST] have a continual [/INST] exchange of what's evolving in which way and different planets. And then each planet would learn from the changes that they see at the other planet. </s><s>[INST] Does the evolution [/INST] happen at the scale of human or </s><s>[INST] do we need the individual like or [/INST] is it more efficient to do like bacteria </s><s>[INST] bacteria [/INST] cheaper and faster and easier. But we but we also have a lot of bacteria in us on us and all around us and even the bacteria in the space station are continually evolving. </s><s>[INST] So did you study that [/INST] by the way like non human cells? </s><s>[INST] Like what the microbiome [/INST] seems, yep. So we've seen that for the astronauts we can actually see </s><s>[INST] there. Their [/INST] immune system respond to the microbiome of the space station. So as soon as you get into that aluminum tube there's a whole </s><s>[INST] ecosystem that's already up there. [/INST] And we can actually see we saw this with scott kelly, </s><s>[INST] we've seen restaurants [/INST] you can see the T </s><s>[INST] cells in their body. They actually [/INST] responding to little peptides </s><s>[INST] a little the [/INST] molecules of the bacteria there there are the </s><s>[INST] immune system is [/INST] looking for a specific bacteria and then once it sees new ones it remembers it </s><s>[INST] and you can see [/INST] the body looking for </s><s>[INST] the microbes that are [/INST] only on the space station </s><s>[INST] that you don't see on [/INST] earth. And when scott came back he actually </s><s>[INST] had more of those microbes [/INST] embedded in his skin in his mouth and stool. They weren't there before. So we're like picked up new </s><s>[INST] Hitchhikers [/INST] in the space station and brought something </s><s>[INST] back down with them. So [/INST] there's like long term </s><s>[INST] ecosystems [/INST] 20 years. They've been there for 20 years. Yes </s><s>[INST] that's there's [/INST] some like chuck Norris </s><s>[INST] that bacteria up [/INST] there I'm sure </s><s>[INST] um [/INST] you you're part </s><s>[INST] of the Extreme microbiome [/INST] project. </s><s>[INST] What does that [/INST] involve? What kind of fun um fun organisms </s><s>[INST] have you [/INST] learned about? Have you </s><s>[INST] gotten to explore? We [/INST] have a really fun project X. M. P. The extreme microbiome. Which is as it sounds like we look for really odd places like heavy radiation environments, high salt high </s><s>[INST] or low temperature you [/INST] know. Strange area the space station for example. Lots of radiation and micro gravity places where organisms can </s><s>[INST] evolved for [/INST] interesting adaptations. And some of them have been organisms. We've seen like a candy pink lake in Australia called Lake Hillier </s><s>[INST] which was just published [/INST] a paper on this. Where </s><s>[INST] is the pink? [/INST] So it's actually done Ilya Salinas. And he's one of these organisms. There's a mixture of bacteria and some allergy there there that make it bright pink. So they actually make </s>

<s>[INST] it seems like this kind of model can be applicable so many aspects like at the different levels of understanding of our reality. So it could be the biology and the chemistry at the physics level. The economics. And you could potentially [/INST] thing is </s><s>[INST] it's like, okay, sure, at all these levels in my rhyme. It might make sense as a model. The question is, can you make useful predictions at these [/INST] levels? That's that's right. And that's really a question of, you know, it's a weird situation because the situation where the model probably has definite consequences. The question is, are there consequences we care about? And that's some uh, you know, and so so in the case of in the economic case, the uh where um so, you know, the 11 thing is this this idea of using kind of physics like notions to construct a kind of distributed analog Blockchain. Okay, the much more pragmatic thing is a different direction and it has to do with this computational language that we built to describe the world that knows about, you know, different kinds of cookies and knows about different cities and knows about how to compute all these kinds of things. One of the things that is of interest is if you want to run the world you need, you know, with with contracts and laws and rules and so on. There are rules at a human level and there are kind of things like, and so this, this gets one into the idea of computational contracts right now, when we write a contract, it's a piece of legalese, it's, you know, it's just written in english and it's not something that's automatically analyze herbal execute double, whatever else, it's just English, you know, back in Gottfried Leibnitz back in, you know, 1680 or whatever was like going to, you know, figure out how to use logic to decide legal cases and so on. And he had kind of this idea of let's make a computational language for the human human law. Forget about modeling nature, forgot about natural laws. What about human law, can we make kind of a computational representation of that? Well, I think finally, we're close to being able to do that. And one of the projects that I hope to get to as soon as the there's a little bit of slowing down of some of this Cambrian explosion that's happening is a project I've been meaning to really do for a long time, which is what I'm calling the symbolic discourse language. It is just finishing the job of being able to represent everything like the conversation we're having in computational terms. And one of the use cases for that is computational contracts. Another use case is something like the constitution that says what the ai is, what we want the ai is </s><s>[INST] to do so, but this is useful. So you're saying uh so these are like you're saying competition contracts. But the smart contracts, this is what in the domain of Cryptocurrency is known as smart contracts. And so the language you've developed this symbolic or seek to further develop symbolic discourse language enables you to uh [/INST] write a contract. So contract </s><s>[INST] that richly represents some aspect of the world, [/INST] but so, I mean, smart contracts tend to be right now, mostly about things happening on the Blockchain, </s><s>[INST] sometimes they have [/INST] oracles and in fact, our Wolfram Alpha api is the main thing people use to get information about the real world, so to speak, within smart contracts. </s><s>[INST] So well from Alpha, as it stands is a really good oracle. Yes, whoever wants to use it, that's perhaps where the relationship with Cardinal. [/INST] Yeah, that's how we started getting involved with Blockchain is we realized people were using, you know, Wolfram Alpha as the oracle for smart contract, so to speak. And so that got us interested in block chains in general. And what was ended up happening is, well from language is with its symbolic representation of things, is really very good at representing things like block chains. And so I think we now have them, we don't really know all the comparisons, but we now have a really nice environment within Wolfram language for dealing with the sort of uh you know, for representing what happens in block chains, for analyzing what happens in block chains. We have a whole effort in Blockchain analytics and, you know, we've we've sort of published some samples of how that works, but it's, you know, because our technology stack Wolfram language in Mathematica very widely used in the quant finance world, there's a there's a sort of immediate uh sort of uh co evolution there of of sort of the quant finance kind of thing. and Blockchain analytics and that's so, so it's kind of the representation of Blockchain in computational language then, ultimately, it's kind of like how do you run the world with code that is, how do you write all these things which are right now, regulations and laws and contracts and things in computational language and kind of the ultimate vision, is that sort of something happens in the world. And then there's this giant domino effect of all these computational contracts, the trigger based on the thing that happened. And there's a there's a whole story to to that. And of course, you know, I I like to always pay attention to the latest things that are going on. And I really, I kind of like Blockchain because it's a it's a it's another rethinking of kind of computation, it's kind of like, you know, cloud computing was a little bit of that of sort of persistent kind of computational resources and so on. And uh you know, this multi computation is a big rethinking of kind of what it means to compute Blockchain is another bit of rethinking of what it means to compute the idea that you lodge kind of these autonomous lumps of computation out there in the Blockchain world. And one of the things that just sort of, for fun, so to speak, as we've been doing a bit of stuff with N. F. T. S. And we just did some N F T s on Gordana and we'll be doing some more and uh, you know, we did some celery or thomason, N F. T s on katana people like, like quite a bit. Um, you know, one of the things I've realized about, about N. F. T. S is that there's kind of this notion and we're really working on this, you know, I like recording stuff. You know, one of the things that's come out of, of kind of my science, I suppose, is this, this History matters type type story of, you know, it's not just the current state, it's the history that matters and I've kind of, I don't think this is such and realizing maybe it's not coincidental that I'm sort of the human who has recorded more about themselves than anybody else. And then I end up with these science results that say history matters, which was not those, those things, I didn't think those were connected, but at </s>

<s>[INST] the flip side of that is because there's a greater, greater gap going from real world to instagram world. You start to be able to laugh at it and realize that it's artificial. So for example, even just artificial filters, people start to realize like there's like it's the same kind of gap as there is between the video game world, in the real world, In the video game world, you can do all kinds of wild things. Uh, Grand theft Auto, you can shoot people up, you can do whatever the heck you want, the real world, you can't and you start to develop an understanding of how to have fun in, in the virtual world and in the physical world. And I think it's just push back. I'm not saying either it's true though. It was a very interesting claims. The more ridiculously out of touch instagram becomes, the easier you can laugh it off potentially in terms of the effect it has on your [/INST] cycle will respond to that. But at some point we should get back to flow. Um, </s><s>[INST] as we engage in flow, [/INST] you laugh at the the hair, the shampoo commercial. You buy the shampoo. </s><s>[INST] Yeah, [/INST] there's a capacity for tremendous bullshitting because of the way these machines are designed to trigger salience without triggering reflective truth seeking. </s><s>[INST] I'm thinking of Connor examples because you're sometimes you're, you can laugh all the way to the bank. So you can, you can laugh and not by the shampoo, right? There's, there's many cases. So I think you have to laugh hard enough. [/INST] You do have to laugh hard enough, but the advertising, the advertisers get millions of dollars precisely because for many, many people, it does make you buy the shampoo and that's the concern </s><s>[INST] and maybe the machine of social media is such the optimizes the shampoo buying. [/INST] Yes, the point I was trying to make is whether or not that particular example is, you know, ultimately right, the possibility of transfer failure is a real thing. Um, and I want to contrast that to an experience I had when I was in grad school and I've been doing tai chi Chuan about three or four years, very religiously in both senses of the word like three or four hours a day and like reading all the literature and right, I was having all the weird experiences, you know, cold as ice, hot as lava, all that stuff and right, but my friends in grad school, they said to me, what's, what's, what's going on? You're different. And I said, what do you mean? And they said, well, you're a lot more balanced in your interactions and you're a lot more flowing and you're a lot more sort of flexible and you adjust more. And I realized, oh, you know, and this was, this was the sort of Taoist claim around tai chi one that it actually transfers in ways that you might not expect you start to be able and I've now noticed that I now notice how, how I'm doing tai chi even in this interaction and how it can facilitate and afford. And so there's a powerful transfer and that's what I meant by, you know, flo wisely not only flow in a way, that's right. Making sure that you're distinguishing causation from correlation which flow can do but find how to situate at home it so that it will percolate through your psyche and permeate through many domains of your life. </s><s>[INST] Is there something you could say similar to our discussion about mindfulness and meditation and contemplation about the world that psychedelics take our mind? Where where does it, where does the mind go when it's on psychedelics? [/INST] I want to remind you of something you said, which is is a gem, it's not so much the experience, but the degree to which it can be integrated back. So, here's the proposal comes from Woodward and others. A lot of convergence around this. Carhartt Harris is talking about it similarly in the entropic brain, but I'm not gonna talk first about psychedelics. I'm gonna talk about neural networks and I'm going to talk about a classic problem in neural networks, or neural networks like us with intuition and implicit learning are fantastic at picking up on complex patterns. Uh </s><s>[INST] talking [/INST] about, I'm talking about a general just general </s><s>[INST] artificial and biological. [/INST] Yes. Yes, Yes. Um, I think at this point there is no relevant difference. So one of the classic problems because of their power is they suffer from overfitting to the data or for those of you are you know, statistical orientation. They pick up patterns in the sample that aren't actually present in the population. Right? And so what you do is there's various strategies you can do drop out where you to periodically turn off half of the nodes in the network, you can drop noise into the network. And what that does is it prevents overfitting to the data and allows the network to generalize more powerfully to the environment. </s><s>[INST] I [/INST] propose to you that that's basically what psychedelics do they, They do that they basically do significant constraint reduction. And so you get areas of the brain talking to each other that don't normally talk to each other, areas that do talk to each other, not talking to each other down regulation of areas that are very dominant, like the default mode network etcetera. And what that does is exactly something strongly analogous sorry to what's happening in drop out or putting noise into the data it opens up by the way, if you give people if you give human beings an insight problem that they're trying to solve and you throw in some noise like literally static on the screen, you can trigger an insight in them. </s><s>[INST] So like literally a very simplistic kind of noise to the perception system [/INST] right? Can break it out of overfitting to the data and open you up now? That means though that just doing that right? In and of itself is not the answer because you also have to make sure that the system can go back to exploring that new space properly. This isn't a problem with neural networks. You turn off drop out and they just go back to being powerful neural networks and now they explore the state space that they couldn't explore before. Human beings are a little bit more messy around this. And this is where the analogy does get a little bit strained. So they need practices to help them integrate that opening up to the new state space so they can properly integrate it. So beyond leary's ST uh set and setting, I think you need another s I think you need sacred, you need psychedelics need to be practiced within a substantial framework, a framework in which people are independently and beforehand improving their abilities to deal with self deception and afford insight and self regulate. This is of course the overwhelming way in which psychedelics are used by indigenous cultures. And I think if we put them into that context, then they can help the project of people self transcending, cultivating meaning and increasing wisdom. But if I think we remove them out of that context and put them in the context of commodities taken just to have certain phenomena logical changes. We run certain important risks. </s>

<s>[INST] Yeah, but you know, that interview process is tricky. I mean, if you're in a rough place mentally, just like with any kind of interview, it's hard not to think that a failed interaction, failed interview, that there's something wrong with you. Sure, right. There is an authority to a therapist, I think, where you think like, they've got it all figured out. Right. And I'm a mess. And therefore, if there's something off, it's all my fault. Right. Right. So it's a very tricky and it's easy to then give up and then because it's like the step to try to get a therapist, the first step to get help, if I get therapist, like any kind of help, [/INST] that's a big leap to take, especially when you're in a rough place. I agree completely. We should not make people swim against such a strong current to get their needs met. I mean, we see this in such obvious places where you have an elderly homebound person who can't get their medicine because, oh, there's been some change and they didn't put the new number into the form or Lord knows what. I mean, it's incredible how we force people to swim against strong currents to get things that are just basic at times for their survival. And with that in mind, I don't have a lot of respect for where healthcare is at or where mental health is at. The field that I work in has accepted all sorts of aspects of how things go, someone else controlling how long the interaction can go on, how the interaction is bounded, what can be said and done, what medicines can be prescribed. There's so many external controls in the systems we work in that we, and I say me included, like all of us in the field have let it get to a place where it's obscenely difficult to get help, obscenely difficult. And we should say that's not okay. I think psychiatrists and therapists and master's level social workers, psychologists, and you name it, I think we should all say this is not okay. And then we as a society should be saying this is not okay. Otherwise, what you're saying, which is I think completely true, will only become worse as there's more and more barriers to getting the help a person needs. And each time a person isn't helped, it sets the odds against them getting more help. I should say here that when I started working, there were times I would send people to an emergency right if there was some emergency emergency in their mental health and they were at risk. And there were times I'd send somebody to an emergency room where if you stopped and looked, it would have been malpractice not to do that, right? Now, it's not just me who has an incredibly high threshold for sending someone to an emergency room because you send someone who's in a lot of distress and oftentimes are sitting on a gurney in a hallway or they're locked in a small white room and all they had was depression, you know, they're scared when they go in and 36 hours later, oh, they're feeling a little better, why? They're desperate to get out of there and someone sends them home. I mean, so our systems have shifted so much that we tolerate now en masse what is </s><s>[INST] egregious to the individual. So you are a psychiatrist in terms of doing therapy, psychotherapy. What does the successful interaction look like? Perhaps a fun question, perhaps not. What do you think of the psychiatrist Sean and Goodwill Hunting played by Robin Williams? So what is the full range of interesting interactions? Can there be an intimacy, a friendship, a kind of varied interaction that kind of blends the lines of, you know, 30 minutes session once a week or whatever versus like a really kind of deliberate long-term project that cares about the well-being of a person across the months and years? What can [/INST] you say about a successful interaction between therapist and patient? I think we're much better served by the latter, right? And again, it doesn't have to be over years. I mean, maybe person doesn't, person might need that over weeks. They might need it over months. They might need it over years. But if I'm understanding correctly, you're describing something that is like a real human engagement. Yeah. Right. And then I work in a field that for years and years and years, the patient didn't get to see, right, the therapist was sitting in a place, sitting behind the person, right? So that's not of course the only tradition and there are aspects of that tradition that can be very humanized. But the idea that we're supposed to not be human...I mean, this medicine is shot through with this, right? That the doctor's supposed to be God and it protects the doctor. And that makes its way into therapy. And the idea of the superiority, the therapist knows more. I mean, in some ways yes, but the idea is like to know more about mechanical things, like right? To know more about about Like facts and knowledge, not as a human being, right? If we approach therapy as a collaborative human endeavor. Right? Or if we're gonna do it together, of course I'm gonna learn from you too. Right? I mean, we're two human beings and we were talking about things that are deep and personal and intimate, and, and I'm don't want, I'm not gonna participate in a way that makes it like about me, as much as it's about you but we're two humans and what's gone on in me may have relevance and sharing it may have relevance. And at times, you doing something back for me may have relevance. I'll give you an example of a person who would not let me help him. It was a young man, so when I was in training, who was very, very sick and needed to change certain choices and habits, or he was not gonna survive. And I had no ability to help him whatsoever. And I went and I saw a supervisor who was existentially trained, where here it's different from existentialism in the classic sense, but it's about really human connection, right? And the guy was always wanting to teach me something, right? Cause I can get by in Spanish, but he was fluent in Spanish. And he wanted to, he'd, oh, you traveled here and he'd say, you know, a word to see if I knew. And I was always directing back to what I was supposed to do, right? And the supervisor, I'll never forget, he said, let him teach you Spanish. Like, okay, come on. So we had a couple of sessions where if you look from the outside, he'd say, what is going on there? Like, right? Like it was, they were Spanish lessons to me, right? And then at some point he brought in his mother and it was he hadn't brought her in yet. And he, he was in part showing off that he taught me something, right? And I said a couple of things and he felt more powerful, like he was younger than me and he felt sick and disempowered, but he didn't feel that way once he taught me something and we showed her off to his mother, right? And his behavior started to change. He started taking better care of himself. He could see a little more what I was saying. He was like, you're a wonderful person. Like, you love your mother and your aunt and they love you. And like, well, you could start seeing that about himself, But that came from humanness. And I think that's the way we help people. I don't understand why we don't do everything that way. It's like, right? We're two humans, but if you're doing something for me then there's something you have an expertise and I don't. That's why you're doing it for me. The reverse could be true, but it doesn't mean we're not </s>

<s>[INST] It's too painful to ask, to raise the taxes to truly transparently pay for the things you're doing with taxes, with taxpayer money, [/INST] because they feel the pain. Either that's one interpretation or it's just too transparent. </s><s>[INST] If people understood the true cost of war, [/INST] they wouldn't wanna go to war. If you were told that you would lose 95% of your assets, and 90% of everything you will be ever, will be taken from you, you might reprioritize your thought about a given policy </s><s>[INST] and you might not vote for that politician. But you're still saying incompetence, not malevolence. So fundamentally, government creates a bureaucracy [/INST] of incompetence is kind of how you look at it. I think a lack of humility, right? Like, if people had more humility, </s><s>[INST] then they would realize. Humility about how little they know, how little they understand [/INST] about the function of complex systems. There's a phrase from Quinn Eastwood's movie, Unforgiven, where he says a man's gotta know his limitations. I think that a lot of people overestimate what they can accomplish and experience. Experience in life causes you to reevaluate that. So, I mean, I've done a lot of things in my life and generally, my mistakes were always my good ideas that I enthusiastically pursued. To the detriment of my great ideas that required 150% of my attention to prosper. So I think people pursue too many good ideas. They all sound good, but there's just a limit to what you can accomplish. And everybody underestimates the challenges of implementing an idea, right? And they always overestimate the benefits of the pursuit of that. And so I think it's an overconfidence that causes an over exuberance in pursuit of policies. And as the ambition of the government expands, so must the currency supply. I could say the money supply, but let's say the currency supply, you can triple the number of pesos in the economy, but it doesn't triple the amount of manufacturing capacity in the said economy. And it doesn't triple the amount of assets in the economy, it just triples the pesos. So as you increase the currency supply, then the price of all those scarce desirable things will tend to go up rapidly. and the confidence of all of the institutions, the corporations and the individual actors </s><s>[INST] and trading partners will collapse. If we take a tangent on a tangent and we will return soon to the big human civilization question, so if government naturally wants to buy stuff it can't afford, what's the best form of government? Anarchism, libertarianism, so not even, there's not even armies, there's no borders, that's anarchism. The least. The smallest possible, the less. [/INST] The best government would be the least </s><s>[INST] and the debate will be over that. When you think about the stuff, do you think about, okay, government is the way it is, I as a person that can generate great ideas, how do I operate in this world? Or do you also think about the big picture, if we start a new civilization somewhere on Mars, do you think about what's the ultimate form of government? [/INST] What's at least a promising thing to try? You know, I have laser eyes on my profile on Twitter, Lex. What does that mean? And the significance of laser eyes is to focus on the thing that can make a difference and if I look at the civilization, I would say half the problems in the civilization are due to the fact that our understanding of economics and money is defective, half, 50%. I don't know, it's worth $500 trillion worth of problems. Like money represents all the economic energy and the civilization and it kind of equates to all the products, all the services and all the assets that we have and we're ever gonna have. So that's half. The other half of the problems in the civilization are medical and military and political and philosophical and natural. And I think that there are a lot of different solutions to all those problems and they're all, they are all honorable professions and they all merit a lifetime of consideration for the specialist in all those areas. I think that what I could offer its constructive is, inflation is completely misunderstood. It's a much bigger problem than we understand it to be. We need to introduce engineering and science techniques into economics if we wanna further the human condition. All government policy is inflationary, you know, and another pernicious myth is, inflation is always and everywhere a monetary phenomena. You know, a famous quote by Milton Friedman, I believe it's like, it's a monetary phenomena that is inflation comes from expanding the currency supply. It's a nice phrase and it's oftentimes quoted by people that are anti-inflation. But again, it just signifies a lack of appreciation of what the issue is. Inflation is, if I had a currency which was completely non-inflationary, if I never printed another dollar and if I eliminated fractional reserve banking from the face of the earth, we'd still have inflation. And we'd have inflation as long as we have government that is capable of pursuing any kind of policies </s><s>[INST] that are in themself inflationary and generally they all are. So in general, inflationary is the big characteristic of human nature that governments collection of groups that have power over others and allocate other people's resources will try to intentionally or not hide the costs of those allocations like in some tricky ways, [/INST] whatever the options are available, tricky ways, whatever. You know, hiding the cost is like the tertiary thing. Like the primary goal is the government will attempt to do good, right? </s><s>[INST] And- [/INST] That's the primary problem? They will attempt to do good and they will do good imperfectly and they will create oftentimes as much damage, more damage than the good they do. Most government policy will be iatrogenic. It will create more harm than good in the pursuit of it but it is what it is. The secondary issue is they will unintentionally pay for it by expanding the currency supply without realizing that they're actually paying for it in a suboptimal fashion. They'll collapse their own currencies while they attempt to do good. The tertiary issue is they will mismeasure how badly they're collapsing the currency. So for example, if you go to the Bureau of Labor Statistics and look at the numbers printed by the Fed, they'll say, oh, it looks like the dollar has lost 95% of its purchasing power over a hundred years. Okay, they sort of fess up there's a problem but they make it 95% loss over a hundred years. What they don't do is realize it's a 99.7% loss over 80 years. So they will miss measure just the horrific extent of the monetary policy in pursuit of the foreign policy and the domestic policy, which they overestimate their budget and their means to accomplish their ends and they underestimate the cost and they're oblivious to the horrific damage that they do to the civilization because the mental models that they use that are conventionally taught are wrong. The mental model that it's okay, we can print all this money because the velocity of the money is low because money velocity is a scalar and inflation is the scalar and we don't see 2% inflation yet and the money velocity is low and so it's okay if we print trillions of dollars. Well, the money velocity was immediate. The velocity of money through the crypto economy 10,000 times faster than the velocity of money through the consumer economy. I think Nick pointed out when you spoke to him, he said it takes two months for a credit card transaction to settle. So you want to spend $1 million in the consumer economy, you can move it six times a year. You put $1 million in a gold, gold will sit in a vault for a decade. So the velocity of money through gold is 0.1. You put the money in the stock market and you can trade it once a week, the settlement is T plus two, maybe you get to two to one leverage. You might get to a money velocity of a hundred a year in the stock market. You put your money into the crypto economy and these people are settling every four hours. And if you're offshore, they're trading with 20x leverage. So if you settle every day and you trade the 20x leverage, you just went to 7,000. Yeah. So the velocity of the money varies. I think the politicians, they don't really understand inflation and they don't understand economics, but you can't blame them because the economists don't understand economics. Because if they did, they would be creating multivariate computer simulations where they actually put in the price of every piece of housing in every city in the world, the full array of foods and the full array of products and the full array of assets. And then on a monthly basis, they would publish all those results and that's a high bandwidth requirement and I think that people don't really want to embrace it. And also there's, the most pernicious thing, there's that phrase, you know, you can't tell people what to think, but you can tell them what to think about. The most pernicious thing is I get you to misunderstand the phenomena so that even when it's happening to you, you don't appreciate that it's a bad thing and you think it's a good thing. So if housing prices are going up 20% year over year and I say, this is great for the American public because most of them are homeowners, then I have misrepresented a phenomena. Inflation is 20%, not 7% and then I have misrepresented it as being a positive rather than a negative and people will stare at it and you could even show them their house on fire and they would perceive it as being great because it's warming them up and they're </s>

<s>[INST] It's kinda interesting that it's not just the length of the podcast medium but it's the fact that it was started by people that don't give a damn about quote unquote demand sort of uh that it's there's a relaxed um sort of the style like that joe Rogan does, there's a freedom to, to um to express ideas in an unconstrained way. That's very real, it's kind of funny in that it feels so refreshingly real to us today and I wonder what the future looks like. It's a little bit sad now that quite a lot of sort of more popular people are getting into podcasting and like, you know, and like and they try to sort of create, they try to control it, they try to constrain it in different kinds of ways. People I love like Conan O'brien and so on different comedians and I'd love to see where the rio aspects of this podcasting medium persists maybe in tv maybe in Youtube, maybe netflix is pushing those kind of ideas and it's kinda, it's really exciting where that kind of sharing of knowledge. [/INST] Yeah, I mean, I think it's a double edged sword as it becomes more popular and more profitable, where on the one hand you'll get a lot more creativity. Um people doing more interesting things with the medium, but also perhaps you get this place to the bottom where suddenly maybe part it will be hard to find good content on podcasts because uh it will be so overwhelmed by, you know, the latest bit of vital outrage. </s><s>[INST] So speaking of that, jumping on effective alternative for a second. Uh huh. So so much of that internet content is funded by advertisements just in your in the context of effective altruism. We're talking about the richest companies in the world, They're funded by advertisements essentially google, that's their primary source of income. Do you see that as do you have any um criticism of that source of income? Do you see that source of money as a potentially powerful source of money that could be used? Well, it certainly could be used for good. But is there something bad about that source of money? [/INST] I think is significant. What is with it? Where it means that the incentives of the company might be quite misaligned with uh you know, making people's lives better where um again, perhaps the incentives are towards increasing drama and debate on your social news, social media feed in order that more people are going to be engaged perhaps kind of uh compulsively involved with the platform, whereas there are other business models, like having an opt in subscription service where perhaps they have other issues, but there's much more of an incentive to provide a product that its users are just like really wanting because you know, now I'm paying for this product, I'm paying for this thing I want to buy rather than I'm trying to use this thing and it's gonna get a profit mechanism that is somewhat orthogonal to me actually just wanting to use the um use the product. And so I mean in some cases it will work better than others. I can imagine a Canon theater. You imagine facebook having a subscription service, but I think it's unlikely to happen anytime soon? </s><s>[INST] Well it's interesting and it's weird now that you bring it up that it's unlikely this example I pay I think $10 a month for YouTube red and that's and I don't think I get it much for that except just uh so no ads but in general it's just a slightly better experience and I would gladly. No, I'm not wealthy. In fact I'm operating very close to $0. Uh but I would pay 10 bucks a month to facebook and 10 bucks a month [/INST] to twitter </s><s>[INST] for um some kind of more control in terms of advertisements and so on. But the other aspect of that is data, personal data, people are really sensitive about this. And I I as one who hopes to one day create a company that may may use people's data to do good for the world. Wonder about this one. The psychology of why people are so paranoid. Well I understand why, but they seem to be more paranoid than is justified at times. And the other is how do you do it? Right? So it seems that facebook as it seems that facebook is doing it wrong. That's certainly the popular narrative, it's unclear to me actually how wrong, like I tend to give them more benefit of the doubt because there, you know, it's a really hard thing to do, right and people don't realize it, but how do we respect in your view, people's privacy. [/INST] Yeah, I mean in the case of how worried are people about using the data? I mean there's a lot of public debate and criticism about it. Um when we look at people's revealed preferences, you know, people's continuing massive use of these sorts of services, uh it's not clear to me how much people really do care. Perhaps they cared a bit, but they're happy to in effect kind of sell the data in order to be able to use a certain service. </s><s>[INST] That's a great term revealed preferences. So these aren't preferences yourself reporting the survey, this is like your actions speak. [/INST] Yeah, exactly. So you might say, oh yeah, I hate the idea of um facebook having my data, but then when it comes to it, you actually are willing to give that data in exchange for um being able to use the service. Uh And if that's the case then I think um unless we have some explanation about why, why there's some negative externality from that or why there's some coordination failure or if there's something that consumers are just really misled about where they don't realize why giving away data like this is a really bad thing to do. Um then ultimately I kind of want to respect people's preferences. They can give away their data if they want. I think there's a big difference between companies use of data and governments having data where you know looking at the track record of history, governments knowing uh a lot about their people um can be very bad if the government chooses to do bad things with it and that's more wedding I think. </s>

<s>[INST] Okay, so interesting. So that that limited participation doesn't require you to run a lot of compute uh does that mean that the richer you are so rich people um are more like their identities more [/INST] right. And </s><s>[INST] this stable, [/INST] Yeah, verifiable </s><s>[INST] or whatever, whatever the right terminology is. [/INST] Right. And this is definitely a copied critique. I think my usual answer to this is that like proof of work is even more of that kind of system is Exactly yeah because </s><s>[INST] I I didn't mean it in that statement is a criticism. I think you're exactly right. That's equivalent that proof of work is the same kind of thing. But in the before Q. Have to also use physical resources. [/INST] Yes. And burn computers and burn trees and all of that stuff. </s><s>[INST] Is there a way to mess with the system of the proof of proof of stake? [/INST] There is but you'll once again need to have a very large portion of all the coins that are locked in the system to do anything bad. Got it. </s><s>[INST] Yeah. And just that maybe take a small change in one of the criticisms of Cryptocurrency is the fact that I guess for the proof of work mechanism have to use so much energy in the world is one of the motivations of proof of stake is to move away from this. Like what's your sense of the uh maybe I'm just under informed. Is there like legitimately environmental impact from this. [/INST] Yeah. So the latest thing was that Bitcoin consumed as much energy as the country of Austria or something like that. And then ethereum is like right now maybe only like half an order of magnitude smaller than Bitcoin. </s><s>[INST] I've heard you talk about a theory into. So what's the What's the dream of a theorem two. So what's the the status of proof of stake because the mechanism that a theory moves forwards and also how do you move to a different mechanism of consensus with within the Cryptocurrency? [/INST] So ethereum 2.0 is a collection of major upgrades that we've wanted to do to ethereum for quite some time. The two big ones, one is a proof of stake and the other is called charting. Um, charting solves another problem with block chains, which is a scalability and what Chardin does is it basically says instead of every participant in the network having to personally download and verify every transaction. Every participant in the network only downwards and verifies a small portion of transactions and you kind of randomly distribute who gets how much work. Um and because of how the distribution is random, it still has the property that you need a large portion of the entire network to corrupt what's going on inside of any shard but the system is still at a very redundant and very secure. </s><s>[INST] That's brilliant. How hard is that to implement and how hard is proof of stake to implement? Like on a technical level, software level, [/INST] proof of stake and charting are both challenging. Might say charting is a bit more challenging. The reason is that proof of stake is kind of just a change to the cow, the consensus lawyer works, charting does both that but it's also a change to the networking layer. Um the reason is that charting is kind of pointless if at the networking layer, you still do what you do today, which is your kind of gossip everything. Which means that if someone publishes something every other note and the client hears it from on the networking layer. And so instead we have to have a sub networks and the ability to quickly switch between sub networks and other sub networks talk to each other. This is all doable, but it's a yeah, more complex architecture and it's definitely the sort of thing that has not yet been done in Cryptocurrency. </s><s>[INST] So most, most of the networking layer and uh Cryptocurrency is you're shouting, you're like broadcast messages and this is more like at hawk networks [/INST] like you're shouting within smaller groups, </s><s>[INST] smaller group. But do you have like a bunch of subjects like exactly, you have to switch between. Oh man, I'd love to see the uh it's a beautiful idea uh from a graph theoretic perspective, but just the software that who's responsible is the theorem project, like the people involved, would they be implementing like, what's the actual, you know, this is like legit software engineering, uh who like how does that work? How do people collaborate build that kind of project? Is this like almost like is there a soft engineering lead? Is there? It's like is a legit almost like large scale open source project. There [/INST] is, Yeah. So we have someone named Danny Ryan on our team who has just been brilliant and great all around and he is a kind of de facto kind of development coordinator, I guess it's like you have to invent job titles for this stuff. And the reason is that like we also have this unique kind of organizational structure where they fear and foundation itself and it does research in the house, but then the actual implementation is done by independent teams that are separate companies and they're located all around the world and like fun places like Australia um and so you know, you kind of just need a bunch of and almost nonstop cat hurting to just keep getting these people to kind of talk to each other and kind of implement this back, make sure that everyone agrees on kind of what's going on and kind of how to interpret different things. </s><s>[INST] So how far into the future are we from these two mechanisms and ethereum 2.0 like what's what's your sense at the time line? Keeping in mind the previous comment you made about the sort of general curse of software projects. [/INST] So Ethereum 2.0, is split into three phases. So phase zero just creates a proof of stake network and it's actually separate from kind of proof of uh the proof of work network at the beginning, just to kind of give it time to grow and improve itself. </s>

<s>[INST] So your book is fascinating. Your perspective on this is fascinating. I'll try to sort of play devil's advocate at times to try to get a clarity. But so the thesis is that war is costly, usually costly for everybody. So that's what you mean when you say nobody wants war because you're going to from a game theoretic perspective, uh, nobody wins. And so war is essentially a breakdown of reason, a breakdown of negotiation of healthy communication or healthy operation of the world. Some kind of breakdown, you list all kinds of ways in which it it breaks down. But there's also there's also human beings in this mix and there is ideas of justice. So for example, I don't want to, my memory doesn't serve me well on which wars were seen as justice. Very, very few in the 20th century of the many that have been there. The wars that were seen as just first of all the most just war is seen as World War Two. By far It's actually the only one that goes above a threshold as seen as just everything is seen as unjust, it's less, it's like degrees of unjust nous and I think the ones that are seen as more just are the ones that are fast that you have a very specific purpose, you communicate that purpose honestly with the global community and you strike hard fast and you pull out uh, to do sort of, it's as a rescue missions, it's almost like policing work. If there's somebody suffering, you go in and stop that suffering directly. And that's it. I think World War Two is seen in that way, that there's an obvious aggressor that is causing a lot of suffering in the world and looking to expand the scale of that suffering. And so you strike, I mean, given the scale, you strike hard as hard and as fast as possible to stop the expansion of the suffering. And so that's kind of how they see. I don't know if you can kind of look with this framework that you've presented and look at Hitler and think, well, it's not in his interest to attack Czechoslovakia Poland Britain France, Russia, the soviet union, America, the United States of America. Same same with Japan. What is it in their interest? Long term interest. I don't know it. Uh, so for me, who cares about alleviating human suffering in the world? Yes, it's not, it seems like almost no words just, but it also seems somehow deeply human to fight. And I think your book makes the case. No, it's not. Can you, can you try to like get at that because it seems that war, there is some that like drum of war seems to beat in all human hearts. Like it's in there somewhere. Maybe it's maybe there's like a relic of the past and we need to get rid of it. It's deeply irrational. Okay [/INST] so obviously we go to war and obviously there's a lot of violence and so we have to explain something and and some of that's going to be aspects of our humanness. So so I guess what I wanted us to sort of start with a I think it was just useful to sort of start and point out actually you know there's really really really really strong incentives not to go to war because it's going to be really costly. And so all of these other human or strategic things. All these things the circumstantial things that will eventually lead us to go to war have to be pretty powerful before we go there. And and most of the time </s><s>[INST] sorry to interrupt. And that's you why you also describe very importantly that war throughout human history is actually rare. We usually avoid it. [/INST] You know most people don't know about The US invasion of Haiti in 1994. I mean I mean a lot of people know about it but people just don't pay attention to it. We don't we we were gonna you know the history books and school kids are going to learn about the invasion of Afghanistan for decades and decades and nobody is going to put this one in the in the history books and it's because it didn't actually happen because before the troops could land the person who had taken power in a coup basically said fine. There's a famous story where where Colin Powell goes to Haiti to this new dictator who has refused to let a democratic president take power and um and tries to convince him to step down or else and he says no no no. And and then he shows them a video and it's basically true planes and all these things taking off and he's like this is not live this is two hours ago. So it's a and and and basically basically gives up right there so that that was </s><s>[INST] that's a powerful move. [/INST] Yeah I think Powell might have been one of his teachers and like a U. S. Military college because a lot of these military dictator trained at some point. And so they had some there was some personal relationships at least between people in the U. S. Government and this guy that they were trying to use. The the point is is and that that's that's like what should have happened like that makes sense right? Like yeah maybe I can mount an insurgency and yeah I'm not going to be a lot of the cost of her because I'm the dictator and maybe he's human and he just wants to fight, gets angry or it's just in his mind whatever he's doing. But at the end of the day is like this does not make sense. Um and that's what happens most of the time, but we don't write so many books about it. And uh and and and now some political scientists go and they count up all of the nations that could fight because they have some dispute and the right next to one another, one another. Or they look at the ethnic groups that could fight with one another because they have there's some tension and they're right next to one another. And And then whatever, some number like 999 out of 1000 don't don't fight because they just find some other way. They don't like each other, but they just loathe in peace because that's the sensible thing to do and that's what we all do. We load in peace. Uh and we loathed the soviet union in relative peace for decades in India. Uh Loath to Pakistan in peace. I mean, two weeks into the Russian invasion of Ukraine. Again, it was in the newspapers, but most people didn't, I think take note India accidentally launched a cruise missile at Pakistan and common suit. So they were like, this is we did not want to go to war. This would be bad. Uh well, x will be angry, but we'll accept your explanation that this was an accident. And so um so these things find on the radar until we overestimate I think how likely to his sides are going to fight. But then of course things do happen like Russia did invade the Ukraine and didn't find some negotiated deal. And so, and so then the book is sort of about half the book is just sort of laying out actually, like there's just different ways this breaks down and some of them are human, some of them are this, I actually don't think war beats in our hearts. It does a little bit. But we're actually very cooperative as a as a species were deeply, deeply cooperative, were really, really good. So those are the things we're not, we're okay at violence and we're okay and we're okay getting angry and vengeance and we have principles that will sometimes lead us, but we're actually really, really, really good at cooperation. And and so that's again, you know, I don't I'm not trying to write some big optimistic book where everything's going to be great and we're all happy and we don't really fight. It's more just to say, let's start, let's be like a doctor as a doctor, we're going to focus on the sick, right? I'm gonna try. I know there's sick people, but I'm going to recognize that the normal state is health and that most people are healthy and and that's going to make me a better doctor. And that's I'm kind of saying the same thing, let's be better doctors of Politics in the world by recognizing that like normal status health. And then we're gonna identify like what are the diseases that are causing this warfare. </s>

<s>[INST] that, that's even bigger than you did. So that's just life. [/INST] That's just, that's just hard. And I think that you can experience so many things. I've had all these injuries. Um, we lost the baby when I was 15, 15 weeks. We've had all these experiences and what the hardest point for me not saying all of those things weren't hard, but it's kind of like you go through that you just realize like life goes on and you have to keep working at it and you have to keep going? And you asked me earlier offline did I feel depressed and um, not from my head injury. I don't think that at least in the moment I had a any recognition of that, it's kind of like, but I think different people's personalities, I have kind of the like buckle down and just keep going. And sometimes it's not until lots of time later that you realize, wow, that was really hard because you were just struggling to live and, and function and do the things that you need to do a </s><s>[INST] long jumping on, just like this part of the conversation just for a few minutes over. Do you mind just sitting together just for a little bit would be [/INST] cool. Be </s><s>[INST] cool. So we put a face to you know uh is it okay with you? That's [/INST] fine with me </s><s>[INST] by the way? What was the head injury if you don't mind sharing? [/INST] Someone had dropped their knee on the back of my head during training was a lot bigger than me. So one strike to the back of the head Too much for someone, there's a reason that's outlawed in Emma, right? Someone 50 lb, everything drops running on the back of your head once. And it's that's the funny thing about getting hit, right? You never can really be sure what's going to happen. I think it's actually one of the magical parts about jujitsu where like if you choke me, if you, we know what's going to occur, you hit someone, they might be completely unharmed, like you might be punching Tony Ferguson in the face and like you need to hit him with a sledgehammer to affect this man, and then other people they could get really badly hurt. Which I guess it's back to your point about street fighting and things like that in a serious, serious potential, you know, 2nd 3rd order consequences of any action that we take. But yeah, that's a that's a tricky thing about getting hit. </s><s>[INST] How does it make you feel that like the really shitty thing about injuries to me, was that like you start thinking like, well if I did this one little thing different, like this wouldn't have happened today, like 11 moment changes your entire life. Is that do you uh do you think that way is that totally counterproductive? [/INST] Um You can't help but think that way when you've had them out of injuries I've had because I've had more than most people's fair share. Um as my orthopedic says, you don't want to win that you don't wanna with the contest of who's had the most to me? A pool. Yeah, but I think you can't help but think that way sometimes, but I definitely don't think it's, I think it can be facilitated. You don't beat yourself up too much because thinking about why have I been subject to so many injuries? And a lot of it comes to just um almost all of mine in particular people a lot heavier than me. So we're but if I've been training martial arts 15 years, I'm obviously on a much smaller side on my woman, I've done thousands and thousands of rounds with people, £50 plus everything. I mean years, not training with anyone less than £50 which is £50 almost half my body weight. And when you also add testosterone in the natural um physiological advantages of men, not just are they heavier with more mass there faster and more explosive, they're stronger um if they're the same size and so I think that the willingness to be in that environment over and over and over again creates a lot of strength, resiliency, willingness to continue. But it also like in order to do that you almost have to uh for me, the way I was approaching it was like pretend like I wasn't more vulnerable um and just be willing to step in and step in and </s><s>[INST] step until you make it until [/INST] you make it kind of Yeah, like I'll just one day I'll be strong enough </s><s>[INST] and you avoided injury for most for most of [/INST] those rounds, I would injury. The problem is Ryan points out is that like you can do thousands of rounds, but if one person that size, that strength that hover reacts in a way that you don't expect, it doesn't, it's not like an oops it's like always major, </s><s>[INST] do you regret any of it? Like [/INST] I think that most no one I know has experienced the degree of injuries that I've experienced and I started using at a time when In 2005 is very different than now where you have the coaches have more control over what you're doing there. More aware in general about a lot of the injuries, there's a lot more people who are uh hobbyists than when I started, there were hobbyists but it was different kind of obvious, you know, then now um now our girls can train with other girls, they don't have to do thousands of rounds with somebody significantly more powerful than them. Um And uh for the drawbacks and the benefits of that, you know, as with anything. Um, so I think, I think that I don't think I would go back and change it. There were times after one of my injuries where I said to Ryan, I said, I quit, I'm done, I'm not doing this anymore. I probably said it more than once, but there was one time I was really serious in 2012 um, I was really serious. I tore my shoulder. I had, I was looking at missing a big competition again the world for my second or third year in a row after injuries and I said I'd quit my job two years before and I'm like, I'm done and run before. That had always been, you know, keep you focused. And then he kind of said, okay, if you want to be done, be done, just just have a good time. No, I'm really done. I didn't want to train anymore. Okay, okay. And then, you know, I think he helped facilitate a moment for me to go um visit a friend, some friend, some girls that were doing a girls camp who are close to my size or some friends of mine to go train and I was like, oh wait, I do love this thing, it was harder for me on a daily basis, but that doesn't mean I don't love this thing and it really helped change my mind. I started to connect with other people travel more myself because previously he had done that, but I hadn't really done that. I think there was a point where um, when I started you too, it was just for fun. I just wanted a sport after college, I played sports as a kid, I want to, I just want to exercise, I wasn't in the martial arts, he used to give me a hard time about it cause he was always very how can you not care about marshall? I don't know, I just want to play sports. Um and Ryan was really big into kind of the philosophy side of the martial arts aspect. Um he used to give me a hard time and I think after that moment, this moment where I looked at myself and I said I want to keep doing this is when I started to appreciate you get to take, it took off some of the pressure had been feeling, I think as Ryan's girlfriend, but I had a full time job a long time, it was never my goal to be a jujitsu world champion and I think after that moment where I was, you know, I really do like this, I really do want to just had this moment, like any time where you're like, I'm doing this for me, I'm not doing this for him and I think that that's um I think that was really lucky for me because how often in our lives do we have a kind of a challenge where we have to stop and we have to say, is this really what I want? How often in a relationship do you do that? How often in any type of lifestyle or job do you stop? And do you really ask yourself if something really difficult happened that you look and you go, am I just doing this because it's convenient and easy or is this what I really want to </s>

<s>[INST] the eye and the AM and the I am is all kind of emerging through this whole process of evolution and so on. That's uh, let's just surface waves and there's a much deeper ocean that we're trying to figure out here. So how does you said they said that some of the stuff you're thinking about maps the spacetime, How does it map the space? [/INST] So, so just a very, very high level and I'll keep it brief. The structures that the physicists are finding Like the Apple two hedren. It turns out they're just static structure, their polyps but they remarkably, most of the information in them is contained in permutation matrices. So it's a matrix, like an N by N matrix that just has zeros and ones that contains almost all of the information. And you can they have these plastic graphs and so forth that they use to boot up the scattering, you can compute the scattering amplitudes almost entirely from these permutation matrices. So that's just now, from my point of view, I have this conscious agent dynamics. It turns out that the stationary dynamics that I was talking about were the entropy is increasing All the stationary dynamics are sketched out by permutation matrices. So, if you so called birkoff polito all the versaces of this public ip all the points, our permutation matrices and all the internal points are marco Vienne kernels um that have the uniform measure as a stationary measure. </s><s>[INST] I need to into it a little better. What what the heck you're talking about. But so basically there's some complicated thing going on with the network of conscience conscious agents and that's map a ble to this. You're saying it two dimensional matrix that uh scattering has to do with what the perception like, that's like photon stuff. I mean, I don't know if it's useful to sort of dig into detail. [/INST] I'll do just a high level thing. So the the high level is the long term behavior of the conscious agent dynamics. So that's the projection of just looking at the long term behavior I'm hoping will give rise to the Apple two hedren. The Upper two Hedren then gives rise to space time. So then I can just use their link to go all the way from consciousness through its asthma topics to through the application into space time and get the map all the way into our interface. </s><s>[INST] And that's why you mentioned the permutation matrix, because it gives you a nice thing to try to generate. It's the [/INST] connection with the amplitude, the permutation matrices are the core of the amplitude Hedren. And it turns out they're the core of the asthma topic, description of the conscious agents. </s><s>[INST] So not to sort of bring up the idea of a creator, but I like, first of all, I like video games and you mentioned this kind of simulation idea first of all, do you think of as an interesting idea, this thought experiment that we live in a simulation. And in general, do you think we live in a simulation? [/INST] So, the nick Bostrom idea about the simulation is typically couched in a physical ist framework. So there's the bottom level there's some programmer in a physical space time and they have a computer that they've programmed, brilliant cleverly where they've created conscious entities. So you have the hard problem of consciousness. Right. The standard heart problem. How could a computer simulation created, which isn't explained by that simulation theory. But then the idea is that the next level the the entities that are from the that are created in the first level simulation then can write their own simulations and you get this this nesting. So so the idea that this is a simulation is fine, but the idea that starts with a physical space, I think, yes, </s><s>[INST] there's there's different properties here. The partial rendering. And to me that's the interesting idea is not whether the entirety of the universe is simulated, but how efficiently can you um create interfaces that are convincing to all other entities that can appreciate such interfaces? How little does it take? Because you said like partial rendering or like temporal, a thermal rendering of stuff? Only render the tree falling in the forest when there's somebody there to see it. It's interesting think how can you do that super efficiently without having to render everything? And that to me is one perspective on the simulation, just like it is with video [/INST] games, where </s><s>[INST] video game doesn't have to render every single thing. It's just the thing that the observer is looking at [/INST] right there is actually that's a very nice question. And there's whole groups of researchers that are actually studying in virtual reality. What what is the sort of minimal requirements on the system. What how does it have to operate to give you an immersion experience to give you the feeling that you have a body to to get you to take it real and and there's actually a lot of really good work on that right now. And it turns out it doesn't take that much. You do need to get the perception action loop tight and you have to give them the perceptions that they're expecting if you want them to, but if you you can lead them along, if you give them perceptions that are close to what they're expecting, you can then maybe move their reality around a bit. </s><s>[INST] Yeah, it's a trick engineering problem, especially when you're trying to create a product that costs little, but that's it feels like an engineering problem, not a deeper scientific problem. Um or meaning obviously it's a scientific problem, but it's a scientific problem. It's not that difficult to trick us descendants of apes. But [/INST] here's here's a case for just us in our own, this is a virtual reality that we're experiencing right now. So here's something you can try for yourself if you just close your eyes and look at we're experience in front of you to be aware of your experience in front of you, which you experience is just like a model dark gray, but there's all sort of there's some dynamics to it, but it's just dark gray. But now I ask you, instead of having your attention forward, put your attention backward, what is it like behind you with your eyes closed? Mhm. And there? it's like nothing. </s>