Gary Marcus: Toward a Hybrid of Deep Learning and Symbolic AI

Gary Marcus is a professor emeritus at NYU, founder of Robust. AI and Geometric Intelligence, the latter is a machine learning company acquired by Uber in 2016. He is the author of several books on natural and artificial intelligence, including his new book Rebooting AI: Building Machines We Can Trust. Gary has been a critical voice highlighting the limits of deep learning and discussing the challenges before the AI community that must be solved in order to achieve artificial general intelligence. This conversation is part of the Artificial Intelligence podcast. If you would like to get more information about this podcast go to https://lexfridman.com/ai or connect with @lexfridman on Twitter https://twitter.com/lexfridman, LinkedIn https://www.linkedin.com/in/lexfridman/, Facebook https://www.facebook.com/lexfridman, Medium https://medium.com/@lexfridman, or YouTube https://www.youtube.com/lexfridman where you can watch the video versions of these conversations. If you enjoy the podcast, please rate it 5 stars on iTunes or support it on Patreon https://www.patreon.com/lexfridman. Here’s the outline with timestamps for this episode (on some players you can click on the timestamp to jump to that point in the episode): 00:00 – Introduction 01:37 – Singularity 05:48 – Physical and psychological knowledge 10:52 – Chess 14:32 – Language vs physical world 17:37 – What does AI look like 100 years from now 21:28 – Flaws of the human mind 25:27 – General intelligence 28:25 – Limits of deep learning 44:41 – Expert systems and symbol manipulation 48:37 – Knowledge representation 52:52 – Increasing compute power 56:27 – How human children learn 57:23 – Innate knowledge and learned knowledge 1:06:43 – Good test of intelligence 1:12:32 – Deep learning and symbol manipulation 1:23:35 – Guitar

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Lex Fridman?

  The following is a conversation with Gary Marcus. He's a professor emeritus at NYU, founder of Robust AI and Geometric Intelligence. The latter is a machine learning company that was acquired by Uber in 2016. He's the author of several books on natural and artificial intelligence, including his new book, Rebooting AI: Building Machines We Can Trust. Gary has been a critical voice highlighting the limits of deep learning and AI in general, and discussing the challenges before our AI community that must be solved in order to achieve artificial general intelligence. As I'm having these conversations, I try to find paths toward insight, towards new ideas. I try to have no ego in the process.

• 0:45Lex Fridman?

  It gets in the way. I'll often continuously try on several hats, several roles. One, for example, is the role of a three-year-old who understands very little about anything and asks big what and why questions. The other might be a role of a devil's advocate who presents counterideas with the goal of arriving at greater understanding through debate. Hopefully, both are useful, interesting, and even entertaining at times. I ask for your patience as I learn to have better conversations. This is the Artificial Intelligence Podcast. If you enjoy it, subscribe on YouTube, give it five stars on iTunes, support it on Patreon, or simply connect with me on Twitter

• 1:28Lex Fridman?

  at LexFriedman, spelled F-R-I-D-M-A-N. And now here's my conversation with Gary Marcus.

• 1:53Lex Fridman?

  Do you think human civilization will one day have to face an AI-driven technological singularity that will, in a societal way, modify our place in the food chain of intelligent living beings on this planet?

• 2:06Speaker 01

  I think our place in the food chain is already changed. So there are lots of things people used to do by hand that they deal with machine. If you think of a singularity as like one single moment, which is, I guess, what it suggests. I don't know if it'll be like that, but I think that there's a lot of gradual change and AI is getting better and better. I mean, I'm here to tell you why I think it's not nearly as good as people think, but you know, the overall trend is clear. Maybe, you know, maybe Ray Kurzweil thinks it's an exponential and I think it's linear. In some cases, it's close to zero right now, but it's all gonna happen. I mean, we are gonna get to human level intelligence or whatever you want, what you will. Artificial general intelligence at some point. And that's certainly gonna change our place in the food chain because a lot of the tedious things that we do now,

• 2:51Speaker 01

  we're gonna have machines do and a lot of the dangerous things that we do now, we're gonna have machines do. And I think our whole lives are gonna change from people finding their meaning through their work, through people finding their meaning through creative expression.

• 3:05Lex Fridman?

  So the the singularity will be a very gradual. In fact, removing the meaning of the word singularity will be a very gradual transformation in your view.

• 3:17Speaker 01

  I think that it'll be somewhere in between. And I guess it depends what you mean by gradual and sudden. I don't think it's gonna be one day. I think it's important to realize that intelligence is a multi-dimensional variable. So, you know, people sort of write this stuff as if like IQ was one number. And, you know, the day that you hit 262 or whatever, you displace the human beings. And really, there's lots of facets to intelligence. So there's verbal intelligence and there's motor intelligence and there's mathematical intelligence and so forth. Machines and their mathematical intelligence far exceed most people already in their ability to play games. They far exceed most people already in their ability to understand language. They lag behind my five-year-old, far behind my five-year-old.

• 4:01Speaker 01

  So there are some facets of intelligence the machines have grasped and some that they haven't. And, you know, we have a lot of work left to do to get them to, say, understand natural language or to understand how to flexibly approach some, you know, kind of novel MacGyver problem solving kind of situation.

• 4:19Speaker 01

  And I don't know that all of these things will come at once. I think there are certain vital prerequisites that we're missing now. So, for example, machines don't really have common sense now. So they don't understand that bottles contain water and that people drink water to quench their thirst and that they don't want to dehydrate. They don't know these basic facts about human beings. And I think that that's a rate-limiting step for many things. It's a rate-limiting step for reading, for example, because stories depend on things like, "Oh, my God, that person's running out of water. That's why they did this thing." Or, you know, "If only they had water, they could put out the fire." So, you know, you watch a movie and your knowledge about how things work matter. And so a computer can't understand that movie if it doesn't have that background knowledge.

• 5:02Speaker 01

  Same thing if you read a book. And so there are lots of places where if we had a good machine interpretable set of common sense, many things would accelerate relatively quickly. But I don't think even that is like a single point. There's many different aspects of knowledge. And we might, for example, find that we make a lot of progress on physical reasoning, getting machines to understand, for example, how keys fit into locks or, you know, that kind of stuff, or, I mean, how this gadget here works and so forth and so on. Machines might do that long before they do really good psychological reasoning. Because it's easier to get kind of labeled data or to do direct experimentation on a microphone

• 5:44Speaker 01

  stand than it is to do direct experimentation on human beings to understand the levers that guide them.

• 5:51Lex Fridman?

  That's a really interesting point, actually, whether it's easier to gain common sense knowledge or psychological knowledge.

• 5:58Speaker 01

  I would say that common sense knowledge includes both physical knowledge and psychological knowledge. And the argument I was making-

• 6:04Lex Fridman?

  Well, you said physical versus psychological.

• 6:06Speaker 01

  Yeah, physical versus psychological. And the argument I was making is physical knowledge might be more accessible because you could have a robot, for example, lift a bottle, try putting a bottle cap on it, see that it, you know, falls off if it does this, and see that it could turn it upside down. And so the robot could do some experimentation. We do some of our psychological reasoning by looking at our own minds. So I can sort of guess how you might react to something based on how I think I I would react to it and robots don't have that intuition and they also can't do experiments on people in the same way or we'll probably shut them down. So, you know, if we wanted to have robots figure out how I respond to pain by pinching me in different ways like that's probably, you know, it's not going to make it past the human subjects board and, you know, companies are going to get sued or whatever.

• 6:49Speaker 01

  So like there's certain kinds of practical experience that are limited or off limits to to robots.

• 6:56Lex Fridman?

  That's a really interesting point. What is more difficult to gain a grounding in? Because to play devil's advocate, I would say that, uh, you know, human behavior is easier expressed in data and digital form. And so when you look at Facebook algorithms, they they get to observe human behavior. So you get to study and manipulate even a human behavior in a way that you perhaps cannot study or manipulate the physical world. So it's true why you said pain is like physical pain, but that's again, the physical world. Emotional pain might be much easier to experiment with, perhaps unethical, but nevertheless, some would argue it's already going on.

• 7:42Speaker 01

  I I think that you're right, for example, that Facebook does a lot of experimentation in in psychological reasoning. In fact, Zuckerberg talked about AI at a talk that he gave NIPS. I wasn't there, but the conference has been renamed NeurIPS, but it's called NIPS when he gave the talk. And he talked about Facebook basically having a gigantic theory of mind. So I think it is certainly possible. I mean, Facebook does some of that. I think they have a really good idea of how to addict people to things. They understand what draws people back to things. I think they exploit it in ways that I'm not very comfortable with. But even so, I think that there are only some slices of human experience that they can access through the kind of interface they have.

• 8:23Speaker 01

  And of course they're doing all kinds of VR stuff and maybe that'll change and they'll expand their data and, you know, I'm sure that that's part of their goal. So it is an interesting question.

• 8:33Lex Fridman?

  I think love. Fear, insecurity, all of the things that I would say some of the deepest things about human nature and the human mind could be explored in the digital form. It's that you're actually the first person just now that brought up, I wonder what is more difficult because I think folks who are the slow, and we'll talk a lot about deep learning, but the people who are thinking beyond deep learning are thinking about the physical world. You're starting to think about robotics in the home robotics. How do we make robots manipulate objects, which requires an understanding of the physical world and requires common sense reasoning? And that has felt to be like the next step for common sense reasoning, but you've not

• 9:17Lex Fridman?

  brought up the idea that there's also the emotional part. And it's interesting whether that's hard or easy.

• 9:23Speaker 01

  I think some parts of it are and some aren't. So my company that I recently founded with Rod Brooks, you know, from MIT for many years and so forth, we're interested in both. We're interested in physical reasoning and psychological reasoning, among many other things. And, you know, there are pieces of each of these that are accessible. So if you want a robot to figure out whether it can fit under a table, that's a relatively accessible piece of physical reasoning. You know, if you know the height of the table and you know the height of the robot, it's not that hard. If you wanted to do physical reasoning about Jenga, it gets a little bit more complicated. You have to have, you know, higher resolution data in order to do it. With psychological reasoning, it's not that hard to know, for example, that people have

• 10:06Speaker 01

  goals and they like to act on those goals, but it's really hard to know exactly what those goals are.

• 10:11Lex Fridman?

  But ideas of frustration, I mean, you could argue it's extremely difficult to understand the sources of human frustration as they're playing Jenga with you or not. You could argue that it's very accessible.

• 10:24Speaker 01

  There's some things that are going to be obvious and some not. So like, I don't think anybody really can do this well yet, but I think it's not inconceivable to imagine machines in the not so distant future, being able to. Understand that if people lose in a game that they don't like that.

• 10:42Lex Fridman?

  Right.

• 10:42Speaker 01

  Right. You know, that's not such a hard thing to program and it's pretty consistent across people. Most people don't enjoy losing and so, you know, that makes it relatively easy to code. On the other hand, if you wanted to capture everything about frustration, well, people get frustrated for a lot of different reasons. They might get sexually frustrated, they might get frustrated they can't get their promotion at work, all kinds of different things. And the more you expand the scope, the harder it is for anything like the existing techniques to really do that.

• 11:09Lex Fridman?

  So I'm talking to Garrett Kasparov next week and he seemed pretty frustrated with his game against Deep Blue, so.

• 11:15Speaker 01

  Yeah, well, I'm frustrated with my game against him last year because I played him. I had two excuses, I'll give you my excuses upfront, but it won't mitigate the outcome. I was jet lagged and I hadn't played in 25 or 30 years. But the outcome is he completely destroyed me and it wasn't even close.

• 11:31Lex Fridman?

  Have you ever been beaten in any board game by a machine?

• 11:36Speaker 01

  I have, I actually beat the, or played the predecessor to Deep Blue. Deep Thought, I believe it was called, and that too crushed me.

• 11:48Lex Fridman?

  And that was, and after that, you realize it's over for us.

• 11:51Speaker 01

  Well, there's no point in my playing Deep Blue, I mean, it's a waste of Deep Blue's computation. I mean, I played Kasparov because we both gave lectures at the same event and he was playing 30 people. I forgot to mention that, not only did he crush me, but he crushed 29 other people at the same time.

• 12:07Lex Fridman?

  I mean, but the actual philosophical and emotional experience of being beaten by a machine, I imagine is a, I mean, to you who thinks about these things, maybe a profound experience. Or no, it was, it was a simple.

• 12:22Speaker 01

  No, I mean.

• 12:23Lex Fridman?

  Mathematical experience?

• 12:24Speaker 01

  Yeah, I think a game like chess particularly where it's, you know, you have perfect information, it's two-player closed end and there's more computation for the computer, it's no surprise the machine wins. I mean, I'm not sad when a computer, I'm not sad when a computer calculates a cube root faster than me. Like, I know I can't win that game, I'm not gonna try.

• 12:45Lex Fridman?

  Well, with a system like AlphaGo or AlphaZero, do you see a little bit more magic in a system like that? Even though it's simply playing a board game, but because there's a strong learning component?

• 12:56Speaker 01

  You know, funny you should mention that in the context of this conversation because Kasparov and I are working on an article that's going to be called AI is Not Magic. And neither one of us thinks that it's magic. And part of the point of this article is that AI is actually a grab bag of different techniques. And some of them have, or they each have their own unique strengths and weaknesses. So you read media accounts and it's like, ooh, AI, it must be magical or can solve any problem. Well, no, some problems are really accessible, like chess and Go. And other problems like reading are completely outside the current technology. And it's not like you can take the technology that drives AlphaGo and apply it to reading and get anywhere. DeepMind has tried that a bit. They have all kinds of resources.

• 13:41Speaker 01

  They built AlphaGo and they have, I wrote a piece recently, they lost, and you can argue about the word lost, but they spent $530 million more than they made last year. So they're making huge investments, they have a large budget. And they have applied the same kinds of techniques to reading or to language. It's just much less productive there because it's a fundamentally different kind of problem. Chess and Go and so forth are closed end problems. The rules haven't changed in 2,500 years. There's only so many moves you can make. You can talk about the exponential as you look at the combinations of moves. But fundamentally, the Go board has 361 squares. That's it. That's the only, those intersections are the only places that you can place your stone.

• 14:23Speaker 01

  Whereas when you're reading, the next sentence could be anything. It's completely up to the writer what they're going to do next.

• 14:31Lex Fridman?

  That's fascinating that you think this way. You're clearly a brilliant mind who points out the emperor has no clothes. But so I'll play the role of a person who says-

• 14:38Speaker 01

  You can put clothes on the emperor? Good luck with it.

• 14:40Lex Fridman?

  Romanticizes the notion of the emperor and period, suggesting that clothes don't even matter. Okay, so that's really interesting that you're talking about language. So there's the physical world of being able to move about the world, making an omelet and coffee and so on. There's language where you-

• 15:01Lex Fridman?

  First, understand what's being written and then maybe even more complicated than that, having a natural dialogue. And then there's the game of Go and chess. I would argue that language is much closer to Go than it is to the physical world. Like, it is still very constrained. When you say the possibility of the number of sentences that could come, it is huge, but nevertheless is much more constrained, it feels, maybe I'm wrong, than the possibilities that the physical world brings us.

• 15:31Speaker 01

  There's something to what you say in some ways in which I disagree. So, one interesting thing about language is that it abstracts away. This bottle, I don't know if it would be in the field of view, is on this table. And I use the word "on" here, and I can use the word "on" here, maybe not here. But that one word encompasses, in analog space, a sort of infinite number of possibilities. So, there is a way in which language filters down the variation of the world. And there's other ways. So, we have a grammar, and more or less, you have to follow the rules of that grammar. You can break them a little bit, but by and large, we follow the rules of grammar. And so, that's a constraint on

• 16:13Speaker 01

  language. So, there are ways in which language is a constrained system. On the other hand, there are many arguments. Let's say there's an infinite number of possible sentences, and you can establish that by just stacking them up. So, I think there's water on the table, you think that I think there's water on the table, your mother thinks that you think that I think the water is on the table, your brother thinks that maybe your mom is wrong to think that you think that I think, right? We can make sentences of infinite length, or we can stack up adjectives. This is a very silly example, a very, very silly example, a very, very, very, very, very silly example, and so forth. There are good arguments that there's an infinite range of sentences. In any case, it's vast by any reasonable measure. And, for example, almost anything in the physical world, we can talk about

• 16:55Speaker 01

  in the language world. And interestingly, many of the sentences that we understand, we can only understand if we have a very rich model of the physical world. So, I don't ultimately want to adjudicate the debate that I think you just set up, but I find it interesting. Maybe the physical world is even more complicated than language. I think that's fair, but-

• 17:16Lex Fridman?

  But you think that language-

• 17:17Speaker 01

  Language is really, really complicated.

• 17:19Lex Fridman?

  It's hard.

• 17:19Speaker 01

  It's really, really hard. Well, it's really, really hard for machines, for linguists, people trying to understand it. It's not that hard for children, and that's part of what's driven my whole career. I was a- Student of Steven Pinker's and we were trying to figure out why kids could learn language when machines couldn't.

• 17:35Lex Fridman?

  I think we're going to get into language, we're going to get into communication intelligence and neural networks and so on, but let me return to the high level, the futuristic, for a brief moment. So you've written in your book, in your new book, it would be arrogant to suppose that we could forecast where AI will be, where the impact will have in a thousand years or even 500 years. So let me ask you to be arrogant. What do AI systems with or without physical bodies look like 100 years from now? If you would just, you can't predict, but if you were to philosophize and imagine, do...

• 18:16Speaker 01

  Can I first justify the arrogance before you try to push me beyond it?

• 18:20Lex Fridman?

  Sure.

• 18:22Speaker 01

  I mean, there are examples, like, you know, people figured out how electricity worked. They had no idea that that was going to lead to cell phones, right? I mean, things can move awfully fast once new technologies are perfected. Even when they made transistors, they weren't really thinking that cell phones would lead to social networking.

• 18:40Lex Fridman?

  There are nevertheless predictions of the future, which are statistically unlikely to come to be, but nevertheless is the best...

• 18:46Speaker 01

  You're asking me to be wrong.

• 18:48Lex Fridman?

  Asking you to be statisti-

• 18:48Speaker 01

  In which way would I like to be wrong?

• 18:50Lex Fridman?

  Pick the least unlikely to be wrong thing, even though it's most, very likely to be wrong.

• 18:56Speaker 01

  I mean, here's some things that we can safely predict, I suppose.

• 18:58Lex Fridman?

  Sure.

• 18:59Speaker 01

  We can predict that AI will be faster than it is now. It will be cheaper than it is now. It will be better in the sense of being more general and applicable in more places. It will be pervasive. You know, I mean, these are easy predictions. I'm sort of modeling them in my head on Jeff Bezos's famous predictions. He says, "I can't predict the future, not in every way." I'm paraphrasing. "But I can predict that people will never want to pay more money for their stuff. They're never going to want it to take longer to get there." And, you know, so like, you can't predict everything, but you can predict some things. Sure, of course, it's going to be faster and better and... What we can't really predict is the full scope of where AI will be in a certain

• 19:44Speaker 01

  period. I think it's safe to say that although I'm very skeptical about current AI, that it's possible to do much better. There's no in-principle argument that says AI is an insolvable problem, that there's magic inside our brains that will never be captured. I've heard people make those kind of arguments. I don't think they're very good. So AI is going to come, and probably 500 years is plenty to get there. And then, once it's here, it really will change everything.

• 20:14Lex Fridman?

  So when you say AI is going to come, are you talking about human-level intelligence?

• 20:20Speaker 01

  I like the term "general intelligence," so I don't think that the ultimate AI, if there is such a thing, is going to look just like humans. I think it's going to do some things that humans do better than current machines, like reason flexibly and understand language and so forth. But that doesn't mean they have to be identical to humans. So, for example, humans have terrible memory, and they suffer from what some people call motivated reasoning. So they like arguments that seem to support them, and they dismiss arguments that they don't like. There's no reason that a machine should ever do that.

• 20:53Lex Fridman?

  So you see that the limitations of memory as a bug, not a feature?

• 21:00Speaker 01

  Absolutely. I'll say two things about that. One is I was on a panel with Danny Kahneman, the Nobel Prize winner, last night, and we were talking about this stuff. And I think what we converged on is that humans are a low bar to exceed. They may be outside of our skill right now, as AI programmers, but eventually, AI will exceed it. So we're not talking about human-level AI. We're talking about general intelligence that can do all kinds of different things and do it without some of the flaws that human beings have. The other thing I'll say is I wrote a whole book, actually, about flaws of humans. It's actually a nice bookend to the, or counterpoint to the current book. So I wrote a book called Cluj, which was about the limits of the human mind. The current book is kind of about those few things that humans do a lot better than machines.

• 21:44Lex Fridman?

  Do you think it's possible that the flaws of the human mind, the limits of memory, our mortality, our bias?

• 21:55Lex Fridman?

  Is a strength, not a weakness, that that is the thing that enables from which motivation springs and meaning springs?

• 22:04Speaker 01

  I've heard a lot of arguments like this. I've never found them that convincing. I think that there's a lot of making lemonade out of lemons. So we, for example, do a lot of free association where one idea just leads to the next and they're not really that well-connected. And we enjoy that and we make poetry out of it and we make kind of movies with free associations and it's fun and whatever. I don't, I don't think that's really a virtue of the system. I think that the limitations in human reasoning actually get us in a lot of trouble. Like, for example, politically, we can't see eye to eye because we have the motivational reasoning I was talking about and something related called confirmation bias. We have all of these problems that actually make for a rougher society because we can't

• 22:46Speaker 01

  get along because we can't interpret the data in, in shared ways. And then we do some nice stuff with that. So my free associations are different from yours and you, you're kind of amused by them and that's great and hence poetry. So there are lots of ways in which we take a lousy situation and make it good. Another example would be our memories are terrible. So we play games like concentration where you flip over two cards, try to find a pair. Can you imagine a computer playing that? Computers like this is the dullest game in the world. I know where all the cards are. I see it once, I know where it is. What are you even talking about? So we make a fun game out of, of having this terrible memory.

• 23:23Lex Fridman?

  So we are imperfect in discovering and optimizing a, some kind of utility function, but you think in general, there is a utility function. There's an objective function that's better than others.

• 23:35Speaker 01

  I didn't say that.

• 23:36Lex Fridman?

  But see, the presumption, when you say.

• 23:41Speaker 01

  I, I think you could design a better memory system. You could argue about utility functions and, and how you want to think about that. But objectively, it would be really nice to do some of the following things to. Get rid of memories that are no longer useful. Like objectively, that would just be good and we're not that good at it. So when you park in the same lot every day, you confuse where you parked today with where you parked yesterday, with where you parked the day before and so forth. So you blur together a series of memories. There's just no way that that's optimal. I mean, I've heard all kinds of wacky arguments, people trying to defend that. But in the end of the day, I don't think any of them hold water.

• 24:17Lex Fridman?

  Or trauma, memories of traumatic events would be possibly a very nice feature to have to get rid of those.

• 24:23Speaker 01

  It'd be great if you could just be like, I'm gonna wipe this sector. You know, I'm done with that. I didn't have fun last night. I don't want to think about it anymore. Whoop, bye bye.

• 24:32Lex Fridman?

  And gone.

• 24:33Speaker 01

  But we can't.

• 24:34Lex Fridman?

  Do you think it's possible to build a system? So you said, human level intelligence is a weird concept, but.

• 24:40Speaker 01

  Well, I'm saying I prefer general intelligence.

• 24:42Lex Fridman?

  General intelligence.

• 24:42Speaker 01

  I mean, human level intelligence is a real thing and you could try to make a machine that matches people or something like that. I'm saying that per se shouldn't be the objective, but rather that we should learn from humans, the things they do well and incorporate that into our AI, just as we incorporate the things that machines do well that people do terribly. So, I mean, it's great that AI systems can do all this brute force computation that people can't. And one of the reasons I work on this stuff is because I would like to see machines solve problems that people can't, that combine the strength or that in order to be solved would combine the strengths of machines to do all this computation with the ability, let's say, of people to read. So, you know, I'd like machines that can read the entire medical literature in a day.

• 25:25Speaker 01

  Seven thousand new papers or whatever's the numbers comes out every day. There's no way for, you know, any doctor or whatever to read them all. A machine that could read would be a brilliant thing and that would be strengths of brute force computation combined with kind of subtlety and understanding medicine that, you know, a good doctor or scientist has.

• 25:43Lex Fridman?

  So if we can linger a little bit on the idea of general intelligence. So Jan LeCun believes that human intelligence isn't general at all. It's very narrow. How do you think?

• 25:53Speaker 01

  I don't think that makes sense. We have lots of narrow intelligences for specific problems. But the fact is, like, anybody can walk into, let's say, a Hollywood movie and reason about the content of almost anything that goes on there. So you can reason about what happens in a bank robbery, or what happens when someone is infertile and wants to, you know, go to IVF to try to have a child. Or you can, you know, the list is essentially endless. And, you know, not everybody understands every scene in a movie, but there's a huge range of things that pretty much any ordinary adult can understand.

• 26:31Lex Fridman?

  His argument is, is that actually, the set of things seems large for us humans because we're very limited in considering the kind of possibilities of experiences that are possible. But in fact, the amount of experience that are possible is infinitely larger than-

• 26:49Speaker 01

  Well, I mean, if you want to make an argument that humans are constrained in what they can understand, I have no, I have no issue with that. I think that's right. But it's still not the same thing at all as saying, here's a system that can play Go. It's been trained on five million games. And then I say, can it play on a rectangular board rather than a square board? And you say, well, if I retrain it from scratch on another five million games, it can. That's really, really narrow, and that's where we are. We don't have even a system that could play Go, and then without further retraining, play on a rectangular board, which any good human could do, you know, with very little problem. So that's what I mean by narrow. And so it's just wordplay to say-

• 27:33Lex Fridman?

  Then it's semantics. Then it's just words. Then, yeah, you mean general in a sense that you can do all kinds of Go board shapes flexibly.

• 27:42Speaker 01

  Well, I mean, that would be like a first step in the right direction. But obviously, that's not what I'm really meaning. You're kidding. What I mean by a general is that you could transfer the knowledge you learn in one domain to another. So if you learn about bank robberies in movies and their chase scenes, then you can understand that amazing scene in Breaking Bad when Walter White has a car chase scene with only one person. He's the only one in it. And you can reflect on how that car chase scene is like all the other car chase scenes you've ever seen, and totally different, and why that's cool.

• 28:17Lex Fridman?

  And the fact that the the number of domains you can do that with is finite doesn't make it less general. So the idea of generals, you could just do it on a lot of transferred across a lot of domains.

• 28:26Speaker 01

  Yeah, I mean, I'm not saying humans are infinitely general or that humans are perfect. I just said, you know, a minute ago, it's a low bar, but it's just, it's a low bar. But, you know, right now, like, the bar is here and we're there and eventually we'll get way past it.

• 28:39Lex Fridman?

  So, speaking of low bars, you've highlighted in your new book as well, but a couple of years ago, wrote a paper titled Deep Learning, A Critical Appraisal, that lists 10 challenges faced by current deep learning systems. So let me summarize them as data efficiency, transfer learning, hierarchical knowledge, open-ended inference, explainability, integrating prior knowledge, cause of reasoning, modeling an unstable world, robustness, adversarial examples, and so on. And then my favorite probably is reliability and engineering of real world systems. So whatever people can read the paper, they should definitely read the paper, should definitely read your book. But which of these challenges, if solved,

• 29:24Lex Fridman?

  in your view, has the biggest impact on the AI community?

• 29:27Speaker 01

  It's a very good question. And I'm going to be evasive because I think that they go together a lot. So some of them might be solved independently of others. But I think a good solution to AI starts by having real, what I would call cognitive models of what's going on. So right now, we have a approach that's dominant where you take statistical approximations of things, but you don't really understand them.

• 29:52Lex Fridman?

  Yes.

• 29:52Speaker 01

  So you know that, you know, bottles are correlated in your data with bottle caps, but you don't understand that there's a thread on the bottle cap that that fits with the thread on the bottle and that that's tightens in if I tighten enough that there's a seal and the water can come out. Like, there's no machine that understands that. And having a good cognitive model of that kind of everyday phenomena is what we call common sense. And if you had that, then a lot of these other things start to fall into at least a little bit better place. Right now, you're like learning correlations between pixels when you play a video game or something like that. And it doesn't work very well. It works when the video game is just the way that you studied it. And then you alter the video game in small ways, like you move the paddle and break out a few pixels, and the system falls apart. Because it doesn't understand, it doesn't have a representation

• 30:36Speaker 01

  of a paddle, a ball, a wall, a set of bricks, and so forth. And so it's reasoning at the wrong level.

• 30:43Lex Fridman?

  So the idea of common sense, it's full of mystery. You've worked on it, but it's nevertheless full of mystery, full of promise. What does, what does common sense mean? What does knowledge mean? So the way you've been discussing it now is very intuitive. It makes a lot of sense that that is something we should have. And that's something deep learning systems don't have. But the argument could be that we're oversimplifying it because we're oversimplifying the notion of common sense, because that's how we, we, it feels like we as humans at the cognitive level approach problems.

• 31:15Speaker 01

  So a lot of people aren't actually going to read my book. But if they did read the book, one of the things that might come as a surprise to them is that we actually say common sense is really hard and really complicated. So they would probably, you know, my critics know that I like common sense, but you know, that chapter actually starts by us beating up not on deep learning, but kind of on our own home team, as it will. So Ernie and I are first and foremost people that believe in at least some of what good old fashioned AI tried to do. So we believe in symbols and logic and programming. Things like that are important. And we go through why even those tools that we hold fairly dear aren't really enough. So we talk about why common sense is actually many things.

• 31:59Speaker 01

  And some of them fit really well with those classical sets of tools. So things like taxonomy. So I know that a bottle is an object or it's a vessel, let's say, and I know a vessel is an object and objects are material things in the physical world. So like, I can make some inferences. If I know that vessels need to, you know, not have holes in them, then I can infer that in order, you know, to carry their contents, then I can infer that a bottle shouldn't have a hole in order to carry its contents. So you can do hierarchical inference and so forth. And we say that's great, but it's only a tiny piece of what you need. For common sense. We give lots of examples that don't fit into that. So another one that we talk about is a cheese grater. You've got holes in a cheese grater.

• 32:44Speaker 01

  You've got a handle on top. You can build a model in the game engine sense of a model so that you could have a little cartoon character flying around through the holes of the grater. But we don't have a system yet. Taxonomy doesn't help us that much. They really understands why the handle is on top and what you do with the handle. Or why all of those circles are sharp. Or how you'd hold the cheese with respect to the grater in order to make it actually work.

• 33:08Lex Fridman?

  Do you think these ideas are just abstractions that could emerge on a system like a very large deep neural network?

• 33:16Speaker 01

  I'm a skeptic that that kind of emergence per se can work. So I think that deep learning might play a role in the systems that do what I want systems to do, but it won't do it by itself. I've never seen a deep learning system really extract an abstract concept. What they do, principled reasons for that, stemming from how backpropagation works, how the architectures are set up. One example is deep learning people actually all build in something like build in something called convolution, which Jan LeCun is famous for, which is an abstraction. They don't have their systems learn this. So the abstraction is an object that looks the same if it appears in different places. And what LeCun figured out and why, you know, essentially why he was a co winner of the Turing Award,

• 34:00Speaker 01

  was that if you programmed this in innately, then your system would be a whole lot more efficient. In principle, this should be learnable. But people don't have systems that kind of reify things and make them more abstract. And so what you'd really wind up with if you don't program that in advance is a system that kind of realizes that this is the same thing as this. But then I take your little clock there and I move it over and it doesn't realize that the same thing applies to the clock.

• 34:27Lex Fridman?

  So the really nice thing, you're right, that convolution is just one of the things that's like, it's an innate feature that's programmed by the human expert.

• 34:35Speaker 01

  We need more of those, not less.

• 34:37Lex Fridman?

  Yes. So the, but the nice feature is it feels like that requires coming up with that brilliant. Idea can get you a Turing Award, but it requires less effort than encoding, and something we'll talk about, the expert system. So encoding a lot of knowledge by hand. So it feels like one, there's a huge amount of limitations, which you clearly outline with deep learning. But the nice feature of deep learning, whatever it is able to accomplish, it does it, it does a lot of stuff automatically without human intervention.

• 35:11Speaker 01

  Well, and that's part of why people love it. Right. But I always think of this quote from Bertrand Russell, which is, it has all the advantages of theft over honest toil. It's really hard to program into a machine, a notion of causality, or even how a bottle works, or what containers are. Ernie Davis and I wrote a, I don't know, 45-page academic paper trying just to understand what a container is, which I don't think anybody ever read the paper. But it's a very detailed analysis of all the things, well, not even all of it, some of the things you need to do in order to understand a container. It would be a whole lot nice. And, you know, I'm a co-author on the paper. I made it a little bit better. But Ernie did the hard work for that particular paper. And it took him like three months

• 35:54Speaker 01

  to get the logical statements correct. And maybe that's not the right way to do it. It's a way to do it. But on that way of doing it, it's really hard work to do something as simple as understanding containers. And nobody wants to do that hard work. Even Ernie didn't want to do that hard work. Everybody would rather just like feed their system in with a bunch of videos with a bunch of containers and have the systems infer how containers work. It would be like so much less effort. Let the machine do the work. And so I understand the impulse. I understand why people want to do that. I just don't think that it works. I've never seen anybody build a system that in a robust way can actually watch videos and predict exactly which containers would leak

• 36:37Speaker 01

  and which ones wouldn't, or something like. And I know someone's going to go out and do that since I said it, and I look forward to seeing it. But getting these things to work robustly is really, really hard. So Jan LeCun, who was my colleague at NYU for many years, thinks that the hard work should go into defining an unsupervised learning algorithm that will watch videos, use the next frame, basically, in order to tell it what's going on. And he thinks that's the royal road, and he's willing to put in the work in devising that algorithm. Then he wants the machine to do the rest. And again, I understand the impulse. My intuition, based on years of watching this stuff and making predictions 20 years ago that still hold, even though there's a lot more computation and so forth,

• 37:23Speaker 01

  is that we actually have to do a different kind of hard work, which is more like building a design specification for what we want the system to do, doing hard engineering work to figure out how we do things like what Jan did for convolution in order to figure out how to encode complex knowledge into the system. The current systems don't have that much knowledge other than convolution, which is, again, this, you know, objects being in different places and having the same perception, I guess I'll say, same appearance.

• 37:53Speaker 01

  People don't want to do that work. They don't see how to naturally fit one with the other.

• 37:58Lex Fridman?

  I think that's, yes, absolutely. But also, on the expert system side, there's a temptation to go too far the other way. So it's just having an expert sort of sit down and encode the description, the framework for what a container is, and then having the system reason the rest. From my view, like, one really exciting possibility is of active learning, where it's continuous interaction between a human and machine. As the machine, there's kind of deep learning type extraction of information from data, patterns and so on, but humans also guiding the learning procedures, guiding the, both the process and the framework of how the machine learns, whatever the task is.

• 38:38Speaker 01

  I was with you with almost everything you said, except the phrase "deep learning." What I think you really want there is a new form of machine learning. So let's remember, deep learning is a particular way of doing machine learning. Most often, it's done with supervised data for perceptual categories. There are other things you can do with deep learning. Some of them quite technical, but the standard use of deep learning is I have a lot of examples and I have labels for them. So here are pictures. This one's the Eiffel Tower. This one's the Sears Tower. This one's the Empire State Building. This one's a cat. This one's a pig and so forth. You just get millions of examples, millions of labels. And deep learning is extremely good at that. It's better than any other solution that

• 39:19Speaker 01

  anybody has devised, but it is not good at representing abstract knowledge. It's not good at representing things like bottles contain liquid and have tops to them and so forth. It's not very good at learning or representing that kind of knowledge. It is an example of having a machine learn something, but it's a machine that learns a particular kind of thing, which is object classification. It's not a particularly good algorithm for learning about the abstractions that govern our world. There may be such a thing. Part of what we counsel in the book is maybe people should be working on devising such things.

• 39:53Lex Fridman?

  So one possibility, just I wonder what you think about it, is that deep neural networks do form abstractions, but they're not accessible to us humans in terms of we can't.

• 40:06Speaker 01

  There's some truth in that.

• 40:07Lex Fridman?

  So is it possible that either current or future neural networks form very high level abstractions which are as powerful as are human abstractions of common sense? We just can't get a hold of them. And so the problem is essentially we need to make them explainable.

• 40:25Speaker 01

  This is an astute question, but I think the answer is at least partly no. One of the kinds of classical neural network architectures that we call an auto-associator. It just tries to take an input, goes through a set of hidden layers and comes out with an output. And it's supposed to learn essentially the identity function, that your input is the same as your output. So you think of those binary numbers, you've got like the one, the two, the four, the eight, the 16 and so forth. And so if you want to input 24, you turn on the 16, you turn on the eight. It's like binary 1, 1 and a bunch of zeros. So I did some experiments in 1998 with the precursors of contemporary deep learning. And what I showed was you could train these networks on all the even numbers and they would never generalize to the odd number.

• 41:11Speaker 01

  A lot of people thought that I was, I don't know, an idiot or faking the experiment or wasn't true or whatever, but it is true that with this class of networks that we had in that day, that they would never, ever make this generalization. And it's not that the networks were stupid, it's that they see the world in a different way than we do. They were basically concerned, what is the probability that the rightmost output node is going to be a one? And as far as they were concerned, in everything they'd ever been trained on, it was a zero. That node had never been turned on, and so they figured, well, I turned it on now. Whereas a person would look at the same problem and say, well, it's obvious, we're just doing the thing that corresponds. The Latin for it is mutatus mutandis, we'll change what needs to be changed.

• 41:55Speaker 01

  And we do this, this is what algebra is. So I can do f of x equals y plus 2, and I can do it for a couple of values. I can tell you if y is 3, then x is 5, and if y is 4, x is 6. And now I can do it with some totally different number, like a million, then you can say, well, obviously it's a million and 2, because you have an algebraic operation that you're applying to a variable. And deep learning systems kind of emulate that, but they don't actually do it. The particular example, you could fudge a solution to that particular problem, but the general form of that problem remains, that what they learn is really correlations between different input and output nodes. And they're complex correlations with multiple nodes involved and so forth, but ultimately they're correlative, they're not structured over these operations over variables.

• 42:39Speaker 01

  Now, someday, people may do a new form of deep learning that incorporates that stuff, and I think it will help a lot. And there's some tentative work on things like differentiable programming right now that fall into that category. But the sort of classic stuff like people use for ImageNet doesn't have it, and you have people like Hinton going around saying, symbol manipulation, like what Marcus, what I advocate is like the gasoline engine, it's obsolete, we should just use this cool electric power that we've got with the deep learning. And that's really destructive, because we really do need to have the gasoline engine stuff that represents, I mean, I don't think it's a good analogy, but we really do need to have the stuff that represents symbols. Yeah.

• 43:21Lex Fridman?

  Hinton as well would would say that we do need to throw out everything and start over. So there there is.

• 43:26Speaker 01

  Yeah, you know, Hinton said that to Axios. And I had a friend who interviewed him and tried to pin him down on what exactly we need to throw. And he was very evasive.

• 43:36Lex Fridman?

  Well, of course, because we can't, if he knew.

• 43:39Speaker 01

  That he'd throw it out himself. But yeah, but I mean, you can't have it both ways. You can't be like, I don't know what to throw out, but I am going to throw out the the symbols. I mean, and not just the symbols, but the variables and the operations of variables. And don't forget the operations of variables, the stuff that I'm endorsing and which, you know, John McCarthy did when he founded AI. That stuff is the stuff that we build most computers out of. There are people now who say we don't need computer programmers anymore. Not quite looking at the statistics of how much computer programmers actually get paid right now. We need lots of computer programmers and most of them, you know, they do a little bit of machine learning, but they still do a lot of code, right? Code where it's like, you know, if the value of X is greater than the value of Y, then do this kind of thing, like conditionals and comparing operations over variables.

• 44:24Speaker 01

  Like, there's this fantasy, you can machine learn anything. There are some things you would never want to machine learn. I would not use a phone operating system that was machine learned. Like, you made a bunch of phone calls and you recorded which packets were transmitted and you just machine learned it, it'd be insane. Or to, you know, build a web browser by taking logs of keystrokes and images, screenshots, and then trying to learn the relation between them. Nobody would ever, no rational person would ever try to build a browser that would make. They would use symbol manipulation, the stuff that I think AI needs to avail itself of in addition to deep learning.

• 44:58Lex Fridman?

  Can can you describe what your view of symbol manipulation in its early days? Can you describe expert systems and where do you think they hit a wall or a set of challenges?

• 45:10Speaker 01

  Sure. So, I mean, first, I just want to clarify, I'm not endorsing expert systems per se. You've been kind of contrasting them. There is a contrast, but that's not the thing that I'm endorsing.

• 45:19Lex Fridman?

  Yes.

• 45:20Speaker 01

  So expert systems try to capture things like medical knowledge with a large set of rules. So if the patient has this symptom and this other symptom, then it is likely that they have this disease. So there are logical rules, and they were symbol manipulating rules of just the sort that I'm talking about.

• 45:37Lex Fridman?

  And they encode a set of knowledge that the experts then put in.

• 45:41Speaker 01

  And very explicitly so. So you'd have somebody interview an expert and then try to turn that stuff into rules.

• 45:48Lex Fridman?

  Yeah.

• 45:48Speaker 01

  And at some level, I'm arguing for rules. But the difference is, what those guys did in the 80s was almost entirely rules, almost entirely handwritten with no machine learning. What a lot of people are doing now is almost entirely one species of machine learning with no rules. And what I'm counseling is actually a hybrid. I'm saying that both of these things have their advantages. So if you're talking about perceptual classification, how do I recognize a bottle? Deep learning is the best tool we've got right now. If you're talking about making inferences about what a bottle does, something closer to the expert systems is probably still the best available alternative. And probably we want something that is better able to handle quantitative and statistical information than those classical systems typically

• 46:30Speaker 01

  were. So we need new technologies that are going to draw some of the strengths of both the expert systems and the deep learning, but are going to find new ways to synthesize them.

• 46:39Lex Fridman?

  How hard do you think it is to add knowledge at the low level? So mine human intellects to add extra information to symbol manipulating systems?

• 46:53Speaker 01

  In some domains, it's not that hard, but it's often really hard, partly because a lot of the things that are important, people wouldn't bother to tell you. So if you pay someone on Amazon Mechanical Turk to tell you stuff about bottles, they probably won't even bother to tell you some of the basic level stuff that's just so obvious to a human being and yet so hard to capture in machines.

• 47:20Speaker 01

  You know, they're going to tell you more exotic things and like they're all well and good, but they're not getting to the root of the problem. So untutored humans aren't very good at knowing, and why should they be? What kind of knowledge the computer system developers actually need? I don't think that that's an irremediable problem. I think it's historically been a problem. People have had crowdsourcing efforts and they don't work that well. There's one at MIT, we're recording this at MIT, called Virtual Home, where, and we talk about this in the book, find the exact example there, but people were asked to do things like describe an exercise routine.

• 48:01Speaker 01

  And the things that the people describe are very low level and don't really capture what's going on. So they're like, go to the room with the television and the weights, turn on the television, press the remote to turn on the television, lift weight, put weight down, whatever. It's like very micro level and it's not telling you what an exercise routine is really about, which is like, I want to fit a certain number of exercises in a certain time period. I want to emphasize these muscles. You want some kind of abstract description. The fact that you happen to press the remote control in this room when you, you know, watch this television isn't really the essence of the exercise routine. But if you just ask people like, what did they do? Then they give you this fine grain.

• 48:43Speaker 01

  And so it takes a level of expertise about how the AI works in order to craft the right kind of knowledge.

• 48:51Lex Fridman?

  So there's this ocean of knowledge that we all operate on. Some of it may not even be conscious, or at least we're not able to communicate it effectively.

• 48:59Speaker 01

  Yeah, most of it we would recognize if somebody said it, if it was true or not, but we wouldn't think to say that it's true or not.

• 49:06Lex Fridman?

  It's a really interesting mathematical property. This ocean has the property that every piece of knowledge in it, we will recognize it as true if we're told, but we're unlikely to retrieve it in the reverse. So that, that interesting property. I would say there's a huge ocean of that knowledge. What's your intuition? Is it accessible to AI systems somehow? Can we?

• 49:32Speaker 01

  Not yet. I mean, most of it is not. Well, I'll give you an asterisk on this in a second, but most of it is not ever been encoded in machine interpretable form. And so, I mean, if you say accessible, there's two meanings of that. One is like, could you build it into a machine? Yes. The other is like, is there some database that we could go, you know, download and stick into our machine?

• 49:55Lex Fridman?

  The first thing, no. Could, could we? Is, what's your intuition?

• 49:57Speaker 01

  I think we could. I, I think it hasn't been done right. You know, the closest, and this is the asterisk, is the CYC, PSYCH, system tried to do this. A lot of logicians worked for Doug Lenett for 30 years on this project. I think they stuck too closely to logic, didn't represent enough about probabilities, tried to hand code it. There are various issues and it hasn't been that successful. That is the closest existing system to trying to encode this.

• 50:27Lex Fridman?

  Why do you think there's not more excitement slash money behind this idea currently?

• 50:33Speaker 01

  There was. People view that project as a failure. I think that they confused the failure of a specific instance that was conceived 30 years ago for the failure of an approach, which they don't do for deep learning. So, you know, in 2010, people had the same attitude towards deep learning. They're like, this stuff doesn't really work. And, you know, all these other algorithms work better and so forth. And then certain key technical advances were made, but mostly it was the advent of graphics processing units that, that changed that. It wasn't even anything foundational in, in the techniques. And there were some new tricks, but mostly it was just more compute and more data, things like ImageNet that didn't exist before, that allowed deep learning. And it could be to work.

• 51:17Speaker 01

  It could be that, you know, PSYCH just needs a few more things or something like PSYCH. But the widespread view is that that just doesn't work. And people are reasoning from a single example. They, they don't do that with deep learning. They don't say, nothing that existed in 2010. There were many, many efforts in in deep learning was really worth anything, right? I mean, really, there's no model from 2010 in in deep learning or the predecessors to the deep learning that has any commercial value whatsoever at this point, right? They're they're all failures. But that doesn't mean that there wasn't anything there. I have a friend, I was getting to know him and he said, he said, I I had a company too. I was talking about I had a new company. He said, I had a company too, and it failed. And I said, well, what did you do? And he said, deep learning.

• 52:02Speaker 01

  And the problem was he did it in 1986 or something like that. And we didn't have the tools then or 1990. We didn't have the tools then, not the algorithms. You know, his algorithms weren't that different from model algorithms. But he didn't have the GPUs to run it fast enough. He didn't have the data. And so it failed. It could be that, you know, symbol manipulation per se with modern amounts of data and compute and maybe some advance in compute for that kind of compute might be great. My my perspective on it is not that we want to resuscitate that stuff per se, but we want to borrow lessons from it, bring together with other things that we've learned.

• 52:40Lex Fridman?

  And it might have an ImageNet moment where it would spark the world's imagination and there'll be an explosion of symbol manipulation efforts.

• 52:48Speaker 01

  Yeah, I think that people at AI2, Paul Allen's AI Institute, are trying to do that. Trying to build data sets that, well, they're not doing it for quite the reason that you say, but they're trying to build data sets that at least spark interest in common sense reasoning.

• 53:02Lex Fridman?

  To create benchmarks that

• 53:03Speaker 01

  Benchmarks for common sense. That's a large part of what the AI2.org is working on right now.

• 53:08Lex Fridman?

  So, speaking of compute, Rich Sutton wrote a blog post titled Bitter Lesson. I don't know if you've read it, but he said that the biggest lesson that can be read from 70 years of AI research is that general methods that leverage computation are ultimately the most effective.

• 53:23Speaker 01

  Do you think But most effective at what? Right? So so they have been most effective for perceptual classification problems and for some reinforcement learning problems. He works on reinforcement.

• 53:35Lex Fridman?

  Well, no, let me push back on that. You're actually absolutely right. But I would also say they would have been most effective generally. Because everything we've done up to Would you argue against that? Is, to me, deep learning is the first thing that has been successful at anything in AI. And you're pointing out that this success is very limited, folks. But has there been something truly successful before deep learning?

• 54:08Speaker 01

  Sure, I mean, I want to make a larger point. But on the narrower point, the, you know, classical AI is used, for example, in doing navigation instructions. You know, it's very successful. Everybody on the planet uses it now, like multiple times a day. That's a measure of success, right? So I don't think classical AI was wildly successful. But there are cases like that. It is used all the time. Nobody even notices them because they're so pervasive.

• 54:40Speaker 01

  So there are some successes for classical AI. I think deep learning has been more successful. But my usual line about this, and I didn't invent it, but I like it a lot, is just because you can build a better ladder doesn't mean you can build a ladder to the moon. So the bitter lesson is if you have a perceptual classification problem, throwing a lot of data at it is better than anything else. But that has not given us any material progress in natural language, understanding, common sense reasoning, like a robot would need to navigate a home. Problems like that, there is no actual progress there.

• 55:16Lex Fridman?

  So flip side of that, if we remove data from the picture, another bitter lesson is that you just have a very simple algorithm and you wait for compute to scale. This doesn't have to be learning. It doesn't have to be deep learning. It doesn't have to be data driven, but just wait for the compute. So my question for you, do you think compute can unlock some of the things of? Either deep learning or symbol manipulation that.

• 55:42Speaker 01

  Sure, but I'll I'll put a, you know, proviso on that. More compute's always better, like, nobody's gonna argue with more compute. It's like having more money.

• 55:51Lex Fridman?

  I mean, there's diminishing returns on more money.

• 55:54Speaker 01

  Exactly, there's diminishing returns on more money, but nobody's gonna argue if you wanna give them more money, right? Except maybe the people who signed the giving pledge, and some of them have a problem. They they've promised to give away more money than they're able to. But the rest of us, you know, if you wanna give me more money, fine.

• 56:09Lex Fridman?

  Say more money, more problems, but okay.

• 56:10Speaker 01

  Um, that's true, too. What I would say to you is, your brain uses like 20 watts, and it does a lot of things that deep learning doesn't do, or the symbol manipulation doesn't do that, that AI just hasn't figured out how to do. So it's an existence proof that you don't need, you know, server resources that are Google scale in order to have an intelligence. You know, I I built, with a lot of help from my wife, two intelligences that are 20 watts each and far exceed anything that anybody else has built at a silicon.

• 56:43Lex Fridman?

  Speaking of those two robots, how, what have you learned about AI from having?

• 56:49Speaker 01

  Well, they're not robots, but.

• 56:51Lex Fridman?

  Uh, sorry, intelligent agents.

• 56:53Speaker 01

  Those two intelligent agents, I've learned a lot by watching my two intelligent agents. I think that what's fundamentally interesting, well, one of the many things that's fundamentally interesting about them is the way that they set their own problems to solve. So my two kids are a year and a half apart, they're both, you know, five and six and a half. They play together all the time, and they're constantly creating new challenges. Like, that's what they do, is they make up games, and they're like, well, what if this, or what if that? Or what if I had this superpower? Or what if, you know, you could walk through this wall? So they're doing these what-if scenarios all the time. And that's how they learn something about the world and and grow their minds, and, you know,

• 57:37Speaker 01

  machines don't really do that.

• 57:39Lex Fridman?

  So that's interesting, and you've talked about this, you've written about it, you've thought about it, nature versus nurture.

• 57:46Lex Fridman?

  So what innate knowledge do you think we're born with? And what do we learn along the way in those early months and years?

• 57:54Speaker 01

  Can I just say how much I like that question? You phrased it just right and almost nobody ever does. Which is what is the innate knowledge and what's learned along the way? So many people dichotomize it and they think it's nature versus nurture. When it is obviously has to be nature and nurture. They have to work together. You can't learn the stuff along the way unless you have some innate stuff. But just because you have the innate stuff, doesn't mean you don't learn anything. And so many people get that wrong, including in the field. Like people think if I work in machine learning, the learning side, I must not be allowed to work on the innate side or that will be cheating. Exactly. People have said that to me. And it's just absurd.

• 58:38Speaker 01

  So thank you.

• 58:40Lex Fridman?

  But you know, you could break that apart more. I've talked to folks who study the development of the brain. And I mean, the growth of the brain in the first few days, in the first few months in the womb, all of that, you know, is that innate? So that process of development from a stem cell to the growth, the central nervous system and so on, to the, the information that's encoded through the long arc of evolution. So all of that comes into play and it's unclear. It's not just whether it's a dichotomy or not. It's, it's where most or where the knowledge is encoded. So what's your intuition about the innate knowledge?

• 59:24Lex Fridman?

  The power of it, what's contained in it? What can we learn from it?

• 59:28Speaker 01

  One of my earlier books was actually trying to understand the biology of this. The book was called The Birth of the Mind. Like, how is it the genes even build innate knowledge? And from the perspective of the conversation we're having today, there's actually two questions. One is what innate knowledge or mechanisms or what have you? People or other animals might be endowed with. I always like showing this video of a baby ibex climbing down a mountain. That baby ibex, you know, a few hours after his birth knows how to climb down a mountain. That means that it knows not consciously something about its own body and physics and 3D geometry and all of this kind of stuff.

• 1:00:04Speaker 01

  So there's one question about like, what does biology give its creatures? And what has evolved in our brains? How is that represented in our brains? The question I thought about in the book, The Birth of the Mind. And then there's a question of what AI should have. And they don't have to be the same. But I would say that, you know, it's a pretty interesting set of things that we are equipped with that allows us to do a lot of interesting things. So I would argue or guess, based on my reading of the developmental psychology literature, which I've also participated in, that children are born with an ocean of space, time, other agents, places, and also this kind of mental algebra that I was describing before. No certain causation, if I didn't just say that.

• 1:00:49Speaker 01

  So at least those kinds of things. They're like frameworks for learning the other things.

• 1:00:55Lex Fridman?

  Are they disjoint in your view? Or is it just somehow all connected? You've talked a lot about language. Is it all kind of connected in some mesh that's language-like? Of understanding concepts altogether?

• 1:01:09Speaker 01

  I don't think we know for people how they're represented in machines. Just don't really do this yet. So I think it's an interesting, open question, both for science and for engineering. Some of it has to be at least interrelated in the way that like the interfaces of a software package have to be able to talk to one another. So the systems that represent space and time can't be totally disjointed because a lot of the things that we reason about are the relations between space and time and cause. So, you know, I put this on and I have expectations about what's going to happen with the bottle cap on top of the bottle. And those span space and time. You know, if the cap is over here, I get a different outcome.

• 1:01:52Speaker 01

  If the timing is different, if I put this here after I move that, then, you know, I get a different outcome. That relates to causality. So obviously, these mechanisms, whatever they are, can certainly communicate with each other.

• 1:02:06Lex Fridman?

  So I think evolution had a significant role to play in the development of this whole kluge, right? How efficient do you think is evolution?

• 1:02:15Speaker 01

  Oh, it's terribly inefficient, except that.

• 1:02:18Lex Fridman?

  Well, can we do better?

• 1:02:20Speaker 01

  Let's come to that in a second. It's inefficient, except that, once it gets a good idea, it runs with it. So it took

• 1:02:30Speaker 01

  I guess a billion years, if I meant roughly a billion years to evolve to a vertebrate brain plan. And once that vertebrate playing plan evolved, it spread everywhere. So fish have it and dogs have it and we have it. We have adaptations of it and specializations of it. But and the same thing with a primate brain plan. So monkeys have it and apes have it and we have it. So, you know, there are additional innovations like color vision and those spread really rapidly. So it takes evolution a long time to get a good idea. But, you know, being anthropomorphic and not literal here. But once it has that idea, so to speak, which cashes out into once a set of genes are in the genome, those genes

• 1:03:15Speaker 01

  spread very rapidly. And they're like subroutines or libraries, I guess the word people might use nowadays or be more familiar with. They're libraries that get used over and over again. So once you have the library for building something with multiple digits, you can use it for a hand, but you can also use it for a foot. You just kind of reuse the library with slightly different parameters. Evolution does a lot of that, which means that the speed over time picks up. So evolution can happen faster because you have bigger and bigger libraries. And what I think has happened in attempts at evolutionary computation is that people start with libraries that are very, very minimal, like almost nothing.

• 1:03:58Speaker 01

  And then, you know, progress is slow and it's hard for someone to get a good PhD thesis out of it and they give up. If we had richer libraries to begin with, if you were evolving from systems that hadn't originate structure to begin with, then things might speed up.

• 1:04:13Lex Fridman?

  Or more PhD students, if the evolutionary process is indeed in a meta way runs away with good ideas, you need to have a lot of ideas, pool of ideas in order for it to discover one that you can run away with. And PhD students representing individual ideas as well.

• 1:04:29Speaker 01

  Yeah, I mean, you could throw a billion PhD students at it.

• 1:04:32Lex Fridman?

  Yeah, the monkeys at typewriters with Shakespeare. Yep.

• 1:04:36Speaker 01

  Well, I mean, those aren't cumulative, right? That's just random. And part of the point that I'm making is that evolution is cumulative. So if you have a billion monkeys independently, you don't really get anywhere. But if you have a billion monkeys, and I think Dawkins made this point originally, or probably other people, Dawkins made it very nice in either Selfish Gene or Blind Watchmaker. If there is some sort of fitness function that can drive you towards something, I guess that's Dawkins' point. And my point, which is a variation on that, is that if the evolution is cumulative, I mean, the related points, then you can start going faster.

• 1:05:12Lex Fridman?

  Do you think something like the process of evolution is required to build intelligent systems?

• 1:05:16Speaker 01

  Not logically. So all the stuff that evolution did, a good engineer might be able to do. So for example, evolution made quadrupeds, which distribute the load across a horizontal surface. A good engineer could come up with that idea. I mean, sometimes good engineers come up with ideas by looking at biology. There's lots of ways to get your ideas. Part of what I'm suggesting is we should look at biology a lot more. We should look at the biology of thought and understanding, and the biology by which creatures intuitively reason about physics or other agents, or like, how do dogs reason about people? Like, they're actually pretty good at it. If we could understand, at my college, we joked dog-nition.

• 1:06:00Speaker 01

  If we could understand dog-nition well and how it was implemented, that might help us with our AI.

• 1:06:06Lex Fridman?

  So do you think, do you think it's possible that the kind of time scale that evolution took is the kind of time scale that will be needed to build intelligent systems? Or can we significantly accelerate that process inside a computer?

• 1:06:22Speaker 01

  I mean, I think the way that we accelerate that process is we borrow from biology.

• 1:06:27Speaker 01

  Not slavishly, but I think we look at how biology has solved problems and we say, does that inspire any engineering solutions here?

• 1:06:34Lex Fridman?

  Try to mimic biological systems and then therefore have a shortcut.

• 1:06:38Speaker 01

  Yeah, I mean, there's a field called biomimicry and people do that for like material science all the time. We should be doing the analog of that for AI. And the analog for that for AI is to look at cognitive science or the cognitive sciences, which is psychology, maybe neuroscience, linguistics and so forth. Look to those for insight.

• 1:06:59Lex Fridman?

  What do you think is a good test of intelligence in your view?

• 1:07:02Speaker 01

  So. I don't think there's one good test. In fact, I try to organize a movement towards something called a touring Olympics. And my hope is that Francois is actually going to take, Francois Charlet is going to take over this. I think he's interested in, I just don't have a place in my busy life at this moment. But the notion is that there'll be many tests and not just one because intelligence is multifaceted. There can't really be a single measure of it because it isn't the single thing. Like just at the crudest level, the SAT has a verbal component and a math component because they're not identical. And Howard Gardner has talked about multiple intelligences, like kinesthetic intelligence and verbal intelligence and so forth.

• 1:07:44Speaker 01

  There are a lot of things that go into intelligence and you know, people can get good at one or the other. I mean, in some sense, like every expert has developed a very specific kind of intelligence. And then there are people that are generalists and you know, I think of myself as a generalist with respect to cognitive science, which doesn't mean I know anything about quantum mechanics, but I know a lot about the different facets of the mind. And you know, there's a kind of intelligence to thinking about intelligence. I like to think that I have some of that, but social intelligence, I'm just okay. You know, there are people that are much better at that than I am.

• 1:08:16Lex Fridman?

  Sure, but what would be really impressive to you? You know, I think the idea of a touring Olympics is really interesting, especially if somebody like Francois is running it. But to you in general, not as a benchmark, but if you saw an AI system being able to accomplish something that would impress the heck out of you, what would that thing be? Would it be natural language conversation?

• 1:08:41Speaker 01

  For me personally. I would like to see a kind of comprehension that relates to what you just said. So I wrote a piece in The New Yorker in, I think, 2015, right after Eugene Guzman, which was a software package, won a version of the Turing test. And the way that it did this is it beat, well, the way you win the Turing test, it's called, win it, is, you know, the Turing test is you fool a person into thinking that a machine is a person, is you're evasive, you pretend to have limitations so you don't have to answer certain questions and so forth. So this particular system pretended to be a 13-year-old boy from Odessa who didn't understand English and was kind of sarcastic and wouldn't answer your questions and so forth.

• 1:09:26Speaker 01

  And so judges got fooled into thinking, you know, briefly with a very little exposure as a 13-year-old boy, and it docked all the questions Turing was actually interested in, which is like, how do you make the machine actually intelligent? So that test itself is not that good. And so in The New Yorker, I proposed an alternative, I guess, and the one that I proposed there was a comprehension test. And I must like Breaking Bad, because I've already given you one Breaking Bad example. In that article, I have one as well, which was like something like, if Walter White, you should be able to watch an episode of Breaking Bad, or maybe you have to watch the whole series to be able to answer the question and say, you know, if Walter White took a hit out on Jesse, why did he do that? Right? So if you could answer kind of arbitrary questions about characters' motivations, I would be really impressed with that.

• 1:10:09Speaker 01

  I mean, you build software to do that, they could watch a film, or there are different versions. And so ultimately, I wrote this up with Praveen Paritosh in a special issue of AI Magazine that basically was about the Turing Olympics. There were like 14 tests proposed. And the one that I was pushing was a comprehension challenge, and Praveen, who's at Google, was trying to figure out like how we'd actually run it, and so we wrote a paper together. And you could have a text version, too, or, you know, you could have an auditory podcast version, you could have a written version. But the point is that you win at this test if you can do, let's say, human level or better than humans. At answering kind of arbitrary questions, you know, why did this person pick up the stone? What were they thinking when they picked up the stone? Were they trying to knock down glass?

• 1:10:52Speaker 01

  And I mean, ideally, these wouldn't be multiple choice either because multiple choice is pretty easily gamed. So if you could have relatively open-ended questions and you can answer why people are doing this stuff, I would be very impressed. And of course, humans can do this, right? If you watch a well-constructed movie and somebody picks up a rock, everybody watching the movie knows why they picked up the rock, right? They all know, oh, my gosh, he's gonna, you know, hit this character or whatever. We have an example in the book about when a whole bunch of people say, I am Spartacus. You know this famous scene? You know, the viewers understand, first of all, that everybody or everybody minus one has to be lying. They can't all be Spartacus.

• 1:11:37Speaker 01

  We have enough common sense knowledge to know they couldn't all have the same name. We know that they're lying and we can infer why they're lying, right? They're lying to protect someone and to protect things they believe in. You get a machine that can do that. They can say, this is why these guys all got up and said, I am Spartacus. I will sit down and say, AI has, you know, has really achieved a lot. Thank you.

• 1:11:58Lex Fridman?

  Without cheating any part of the system.

• 1:12:00Speaker 01

  Yeah, I mean, if you do it, there are lots of ways you could cheat. Like, you could build a Spartacus machine that works on that film. Like, that's not what I'm talking about. I'm talking about you can do this with essentially arbitrary films or, you know, from a large.

• 1:12:12Lex Fridman?

  Even beyond films because it's possible such a system would discover that the number of narrative arcs in film is like limited to like.

• 1:12:20Speaker 01

  Well, there's a famous thing about the classic seven plots or whatever. I don't care if you want to build in the system, you know, boy meets girl, boy loses girl, boy finds girl. That's fine. I don't mind having some head start.

• 1:12:30Lex Fridman?

  Innate knowledge?

• 1:12:31Speaker 01

  Okay.

• 1:12:31Lex Fridman?

  Good.

• 1:12:32Speaker 01

  I mean, you could build it in innately or you could have your system watch a lot of films again. If you can do this at all, but with a wide range of films, not just one film in one genre.

• 1:12:44Speaker 01

  But even if you could do it for all Westerns, I'd be reasonably impressed.

• 1:12:46Lex Fridman?

  Yeah. So in terms of being impressed, just for the fun of it, because you've put so many interesting ideas out there in your book, challenging the community for further steps. Is it possible on the deep learning front that you're wrong about its limitations? That deep learning will unlock, Yan Le Cun next year will publish a paper that achieves this comprehension. So do you think that way often as a scientist? Do you consider that your intuition, that deep learning could actually run away with it?

• 1:13:23Speaker 01

  I'm more worried about rebranding is a kind of political thing. So, I mean, what's going to happen, I think, is that deep learning is going to start to encompass symbol manipulation. So I think Hinton's just wrong. You know, Hinton says we don't want hybrids. I think people will work towards hybrids and they will relabel their hybrids as deep learning. We've already seen some of that. So AlphaGo is often described as a deep learning system, but it's more correctly described as a system that has deep learning, but also Monte Carlo tree search, which is a classical AI technique. And people will start to blur the lines in the way that IBM blurred Watson. First Watson meant this particular system, and then it was just anything that IBM built in their cognitive vision.

• 1:14:00Lex Fridman?

  Sure, but purely, let me ask for sure. That's a branding question, and that's a giant mess. I mean, purely, a single neural network being able to accomplish.

• 1:14:10Speaker 01

  I don't, I don't stay up at night worrying that that's going to happen. And I'll just give you two examples. One is a guy at DeepMind thought he had finally outfoxed me. @Zurgielord, I think, is his Twitter handle. And he said, he specifically made an example. Marcus said that such and such, he fed it into GPT-2, which is the AI system that is so smart that OpenAI couldn't release it because it would destroy the world, right? You remember that a few months ago? So he feeds it into GPT-2, and my example was something like a rose is a rose, a tulip is a tulip, a lily is a blank. And he got it to actually do that, which was a little bit impressive. And I wrote back and I said, that's impressive, but can I ask you a few questions?

• 1:14:54Speaker 01

  I said, can it, you know, was that just one example? Can it do it generally? And can it do it with novel words, which was part of what I was talking about in 1998 when I first raised the example? So a DAX is a DAX, right?

• 1:15:07Speaker 01

  And he sheepishly wrote back about 20 minutes later. And the answer was, well, it had some problems with those. So, you know, I made some predictions 21 years ago that still hold. And in the world of computer science, that's amazing, right? Because, you know, there's a thousand or a million times more memory and, you know, computations a million times, you know, do million times more operations per second, you know, spread across a cluster. And there's been advances in, you know, replacing sigmois with other functions and so forth. There's all kinds of advances, but the fundamental architecture hasn't changed and the fundamental limit hasn't changed. And what I said then is kind of still true. Then here's a second example. I recently had a piece in Wired that's adapted from the book.

• 1:15:51Speaker 01

  And the book didn't, it was, went to press before GP2 came out. But we describe this children's story and all the inferences that you make in this story about a boy finding a lost wallet. And for fun, in the Wired piece, we ran it through GP2. GP2 at something called talk2transformer.com and your viewers can try this experiment themselves. Go to the Wired piece that has the link and it has the story. And the system made perfectly fluent text that was totally inconsistent with the conceptual underpinnings of the story. Right? This is what, you know, again, I predicted in 1998. And for that matter, Chomsky Miller made the same prediction in 1963. I was just updating their claim for a slightly new text.

• 1:16:36Speaker 01

  So those particular architectures that don't have any built in knowledge, they're basically just a bunch of layers doing correlational stuff, they're not going to solve these problems.

• 1:16:47Lex Fridman?

  So 20 years ago, you said, "The emperor has no clothes." Today, "The emperor still has no clothes."

• 1:16:53Speaker 01

  The lighting's better, though.

• 1:16:54Lex Fridman?

  The lighting is better. And I think you yourself are also, I mean...

• 1:16:59Speaker 01

  And we found out some things to do with naked emperors. I mean, it's not that like the stuff is worthless. I mean, they're not really naked. It's more like they're in their briefs and everybody thinks that... And so like, I mean, they are great at speech recognition, but the problems that I said were hard. Because I didn't literally say, "The emperor has no clothes." I said, "This is a set of problems." Problems that humans are really good at. And it wasn't couched as AI, it was couched as cognitive science. But I said, if you want to build a neural model of how humans do certain class of things, you're going to have to change the architecture. And I stand by those claims.

• 1:17:30Lex Fridman?

  So, and I think people should understand you're quite entertaining in your cynicism, but you're also very optimistic and a dreamer about the future of AI too. So you're both, it's just.

• 1:17:42Speaker 01

  There's a famous saying about being, you know, people overselling technology in the short run and underselling it in the long run. And so I actually end the book, or Ernie Davis and I end our book with an optimistic chapter, which kind of killed Ernie because he's even more pessimistic than I am. He describes me as a contrarian and him as a pessimist. But I persuaded him that we should end the book with a look at what would happen if AI really did incorporate, for example, the common sense reasoning and the nativism and so forth, the things that we counseled for. And we wrote it, and it's an optimistic chapter. That AI suitably reconstructed so that we could trust it, which we can't now, could really be world changing.

• 1:18:26Lex Fridman?

  So on that point, if you look at the future trajectories of AI, people have worries about negative effects of AI, whether it's at the large existential scale or smaller short term scale of negative impact on society. So you write about trustworthy AI. How can we, how can we build AI systems that align with our values, that make for a better world, that we can interact with, that we can trust?

• 1:18:51Speaker 01

  The first thing we have to do is to replace deep learning with deep understanding. So we need, you can't have alignment with a system that traffics only in correlations and doesn't understand concepts like bottles or harm. So Asmau talked about these famous laws. And the first one was first do no harm. And you can quibble about the details of Asmau's laws, but we have to, if we're going to build real robots in the real world, have something like that. That means we have to program in a notion that's at least something like harm. That means we have to have these more abstract ideas that deep learning is not particularly good at. They have to be in the mix somewhere. And you could do statistical analysis about probabilities of given harms or whatever, but you have to know what a harm is in the same way that you have to understand that a bottle isn't just a collection of pixels.

• 1:19:37Lex Fridman?

  And also be able to, you're implying that you need to also be able to communicate that to humans. So the AI systems would be able to prove to humans that they understand, that they know what harm means.

• 1:19:52Speaker 01

  I might run it in the reverse direction, but roughly speaking, I agree with you. So we probably need to have committees of wise people, ethicists and so forth. Think about what these rules ought to be, and we shouldn't just leave it to software engineers. It shouldn't just be software engineers, and it shouldn't just be, you know, people who own large mega corporations that are good at technology. Ethicists and so forth should be involved. But, you know, there should be some assembly of wise people, as I was putting it, that tries to figure out what the rules ought to be, and those have to get translated into code. You can argue, or code or neural networks or something, they have to be translated into something that machines can work with. And that means there

• 1:20:37Speaker 01

  has to be a way of working the translation. And right now, we don't. We don't have a way. So let's say you and I were the committee, and we decide that Asimov's First Law is actually right. And let's say it's not just two white guys, which would be kind of unfortunate that we have a broad, and so we've representative sample of the world, or however we want to do this. And the committee decides eventually, okay, Asimov's First Law is actually pretty good. There are these exceptions to it. We want to program in these exceptions. But let's start with just the first one, and then we'll get to the exceptions. First one is first do no harm. Well, somebody has to now actually turn that into a computer program or a neural network or something. And one way of taking the whole book, the whole argument that I'm making, is that we just don't know how to do that yet. And we're fooling ourselves

• 1:21:20Speaker 01

  if we think that we can build trustworthy AI, if we can't even specify in any kind of, you know, we can't do it in Python, and we can't do it in TensorFlow. We're fooling ourselves in thinking that we can make trustworthy AI if we can't translate harm into something that we can execute. And if we can't, then we should be thinking really hard, how could we ever do such a thing? Because if we're going to use AI in the ways that we want to use it, to make job interviews, or to do surveillance, not that I personally want to do that, or whatever, I mean, if we're going to use AI in ways that have practical impact on people's lives, or medicine. It's gotta be able to understand stuff like that.

• 1:21:57Lex Fridman?

  So one of the things your book highlights is that, you know, a lot of people in the deep learning community, but also the general public, politicians, just people in all general groups and walks of life have a different levels of misunderstanding of AI. So when you talk about committees, what's your advice to our society? How do we grow? How do we learn about AI such that such committees could emerge? Where large groups of people could have a productive discourse about how to build successful AI systems?

• 1:22:33Speaker 01

  Part of the reason we wrote the book was to try to inform those committees. So part of the reason we wrote the book was to inspire a future generation of students to solve what we think are the important problems. So a lot of the book is trying to pinpoint what we think are the hard problems where we think effort would most be rewarded. And part of it is to try to train people who talk about AI but aren't experts in the field to understand what's realistic and what's not. One of my favorite parts in the book is the six questions you should ask anytime you read a media account. So like, number one is if somebody talks about something, look for the demo. If there's no demo, don't believe it. Like the demo that you can try. If you can't try it at home, maybe it doesn't really work that well yet.

• 1:23:15Speaker 01

  So if, we don't have this example in the book, but if Sundar Pichai says, "We have this thing that allows it to sound like human beings in conversation," you should ask, "Can I try it?" And you should ask how general it is. And it turns out at that time, I'm alluding to Google Duplex when it was announced, it only worked on calling hairdressers, restaurants, and finding opening hours. That's not very general. That's, that's narrow AI.

• 1:23:38Lex Fridman?

  And I'm not going to ask your thoughts about Sophia, but yeah. I understand that's a really good question to ask of any kind of hightop idea.

• 1:23:46Speaker 01

  Sophia has very good material written for her, but she doesn't understand the things that she's saying.

• 1:23:52Lex Fridman?

  So a while ago, you've written a book on the science of learning, which I think is fascinating, but the learning case studies of playing guitar.

• 1:24:00Speaker 01

  That's right.

• 1:24:00Lex Fridman?

  Called Guitar Zero. I love guitar myself. I've been playing my whole life. So let me ask a very important question. What is your favorite song, rock song, to listen to or try to play?

• 1:24:12Speaker 01

  Well, those would be different, but I'll say that my favorite rock song to listen to is probably All Along the Watchtower, the Jimi Hendrix version.

• 1:24:18Lex Fridman?

  The Jimi Hendrix version.

• 1:24:19Speaker 01

  It feels magic to me.

• 1:24:21Lex Fridman?

  I've actually recently learned it. I love that song. I've been trying to put it on YouTube myself, singing. Singing is the scary part. If you could party with a rock star for a weekend, living or dead, who would you choose?

• 1:24:34Lex Fridman?

  And pick their mind. It's not necessarily about the partying.

• 1:24:37Speaker 01

  Thanks for the clarification. I guess John Lennon is such an intriguing person. I mean, think a troubled person, but an intriguing one, so.

• 1:24:48Lex Fridman?

  Beautiful. Well, Imagine is one of my favorite songs, so.

• 1:24:52Speaker 01

  Also one of my favorite songs.

• 1:24:53Lex Fridman?

  That's a beautiful way to end it. Gary, thank you so much for talking today.

• 1:24:56Speaker 01

  Thanks so much for having me.