The following is a conversation with Demis Hassabis, his second time on the podcast. He is the leader of Google DeepMind and is now a Nobel Prize winner. Demis is one of the most brilliant and fascinating minds in the world today working on understanding and building intelligence and exploring the big mysteries of our universe. This was truly an honor and a pleasure for me. And now a quick few second mention of a sponsor. Check them out in the description or at lexfridman.com/sponsors.
It's the best way to support this podcast. We got Hampton for connecting with founders and CEOs, Finn for AI customer service, Shopify for building e-commerce businesses, Element for daily electrolytes, and AG1 for your health. Choose wisely, my friends. And now on to the full ad reads. I do try to make them interesting, but if you must skip, friends, please still check out our sponsors. I enjoy their stuff, maybe you will, too. And also, to get in touch with me for whatever reason, go to lexfridman.com/contact. All right, let's go.
This episode is brought to you by Hampton, a private community for high-growth founders and CEOs. That's the interesting thing about starting a company and running a company, especially one that's growing really quickly, has to hire a lot, has to scale a lot. It's perhaps a little bit counterintuitive, but for the founder, it can be deeply lonely. I suppose that's one of the reasons they recommend to have a co-founder, but even outside of that, there's just a deep loneliness with putting it all on the line, risking everything, knowing that the chances of success are low. But if you do succeed, the gains are huge, and you have your heart in it, you have your dreams in it, you believe in it, but also there's a constant roller coaster of fear and doubt and hope and moments of triumph, and moments of failure. All those go back and forth and just this is a constant psychological turmoil. Anyway, through all that, it's just nice to connect with other people that are going through the same thing, and that's what Hampton is about. It does a thing where every month eight founders face-to-face have real conversations about their daily struggles. Groups are forming in a bunch of places, New York City, Austin, San Francisco, LA, Miami, Denver, and so on. If you are a founder who's tired of carrying it all alone, visit joinhampton.com/lex to see if it's a fit for you. That's joinhampton.com/lex.
This episode is also brought to you by Finn. It's an AI agent for customer service. So, they are focused, laser-focused on the customer service application, and they want to do that better than anybody else in the world. In fact, if you measure by the metric of the number of resolutions, so when you have the agent resolve a customer service issue, that's resolution, they have a 59% average resolution rate, which makes it the highest-performing customer service agent on the market. It's trusted by over 5,000 customer service leaders, and even top AI companies including Anthropic. The way they design the system is it can continuously improve from the interactions, so you can continuously analyze, train, test, and deploy. Also, probably important to say, they give you a 90-day money-back guarantee. Go to fin.ai/lex to learn more about transforming your customer service and scaling your support team. That's fin.ai/lex.
This episode is also brought to you by Shopify, a platform designed for anyone to sell anywhere with a great-looking online store. Even I figured out how to create an online store at lexfridman.com/store. I put up a few shirts. Haven't done anything with it since because I'm not a serious person. There's a lot of serious people that build real businesses on top of Shopify. It's a platform that connects you with millions of people that want to buy stuff and gives you all the tools you need and all the integrations that you need to do just that at scale. As we talked about with DHH about the incredible beauty and power of Ruby on Rails that Shopify is powered by. I have not yet built a serious sort of medium-scale project on Rails. I need to. It's just I need to actually find things that I need to do web dev type of stuff with to inspire myself to build something useful. I don't want to build some weird variant of a to-do list, especially now with the help of LLMs, you can generate so much of the code. So, I need to figure out how to learn new framework and new programming languages when LLMs can generate so much of it. And I don't want to do it exclusively by vibe coding because I feel like that's not a way to learn fully a thing, but vibe coding does remove some of the friction of learning. So, balancing that out is a tricky thing to do. Anyway, that's about the programming language and the framework that powers Shopify. But Shopify itself connects buyers and sellers at an incredible scale that's awe-inspiring. Sign up for a $1 per month trial period at shopify.com/lex. That's all lowercase. Go to shopify.com/lex to take your business to the next level today.
This episode is also brought to you by Element, my daily zero sugar and delicious electrolyte mix. I've been traveling recently and I have a lot of Element packets with me and I bring that and I bring bands, you call them. I don't know what they're called. They're like rubber bands for basic shoulder exercises. So, if I have to do a lot of either heavy lifting or heavy jujitsu training, I like to warm up the shoulders really well because probably I have issues with shoulders from many years of playing tennis and many years of doing stupidly bench press. Anyway, I think Element is a critical component of my workout routine. Hydrate before, rehydrate after, fully embrace the deliciousness of watermelon salt flavor. The flavor of champions, the one I recommend. It's been quite a while since I tried the others. They're all good, but for me, I'm a man of focus and dedication. And I'm dedicated to watermelon salt. I think they have actually, I saw a lemonade flavor. I think a lot of people love lemonade. So, maybe that's your thing. For me, I'm sticking to watermelon salt. Get a free 8-count sample pack with any purchase. Try it at drinkelement.com/lex.
This episode was also brought to you by AG1, an all-in-one daily drink to support better health and peak performance. I travel with it. It makes me feel like I take a little piece of home with me. I drink it at least once a day, very often twice a day. And they keep innovating, they keep improving it. They recently introduced AG1 next gen, improving every aspect, more vitamins and minerals, and upgraded probiotics. It's funny how a morning routine can be the source of peace and happiness. Because I find that if I check my phone at all in the first couple hours of the day, I get this weird anxiety that ultimately morphs into unhappiness. And if I don't, I'm much more likely to sort of maintain that deep focus. And a part of that early in the morning is some coffee or caffeinated drink, and then a few hours on is AG1. And it's just many hours of deep focus in between. It makes me feel happy, makes me feel at one with the universe, and it helps me get done. Anyway, they'll give you a 1-month supply of fish oil when you sign up at drinkAG1.com/lex.
This is a Lex Fridman podcast. To support it, please check out our sponsors in the description or at lexfridman.com/sponsors, and consider subscribing, commenting, and sharing the podcast with folks who might find it interesting. I promise to work extremely hard to always bring you nuanced and long-form conversations with a wide variety of interesting people from all walks of life. And now, dear friends, here's Demis Hassabis.
In your Nobel Prize lecture, you proposed what I think is a super interesting conjecture that, quote, any pattern that can be generated or found in nature can be efficiently discovered and modeled by a classical learning algorithm. What kind of patterns or systems might it be included in that? Biology, chemistry, physics, maybe cosmology,
neuroscience, what are we talking about?
Sure. Well, look, I felt that it's sort of a tradition, I think, of Nobel Prize lectures that you're supposed to be a little bit provocative, and I wanted to follow that tradition. What I was talking about there is if you take a step back and you look at all the work that we've done, especially with the Alpha X projects, so I'm thinking Alpha Go, of course, AlphaFold. What they really are is we're building models of very combinatorially high-dimensional spaces that, if you try to brute-force a solution, find the best move in Go, or find the exact shape of a protein, and if you enumerated all the possibilities, there wouldn't be enough time in the time of the universe. So, you have to do something much smarter. And what we did in both cases was build models of those environments. And that guided the search in a smart way. And that makes it tractable. So, if you think about protein folding, which is obviously a natural system, why should that be possible? How does physics do that? Proteins fold in milliseconds in our bodies. So, somehow physics solved this problem that we've now also solved computationally. And I think the reason that's possible is that in nature, natural systems have structure because they were subject to evolutionary processes that shaped them. And if that's true, then you can maybe learn what that structure is.
So, this perspective, I think, is a really interesting one. You've hinted at it, which is almost like crudely stated, anything that can be evolved can be efficiently modeled. Think there's some truth to that?
Yeah, I sometimes call it survival of the stabilist or something like that because it's of course there's evolution for life, living things, but there's also, if you think about geological time, so the shape of mountains, that's been shaped by weathering processes but over thousands of years. But then you can even take it cosmological, the orbits of planets, the shapes of asteroids. These have all survived kind of processes that have acted on them many many times. So, if that's true, then there should be some sort of pattern that you can kind of reverse learn and a kind of manifold really that helps you search to the right solution, to the right shape, and actually allow you to predict things about it in an efficient way. Because it's not a random pattern. So, it may not be possible for man-made things or abstract things like factorizing large numbers because unless there's patterns in the number space, which there might be, but if there's not and it's uniform, then there's no pattern to learn. There's no model to learn that will help you search. So, you have to do brute force. So, in that case, you maybe need a quantum computer, something like this. But, in most things in nature that we're interested in are not like that. They have structure that evolved for a reason and survived over time. And if that's true, I think that's potentially learnable by a neural network.
It's like nature's doing a search process. And it's so fascinating that in that search process it's creating systems that could be efficiently modeled.
So, they can be efficiently rediscovered or recovered because nature's not random. These everything that we see around us, including like the elements that are more stable, all of those things, they're subject to some kind of selection process pressure.
Do you think, because you're also a fan of theoretical computer science and complexity, do you think we can come up with a kind of complexity class, like a complexity zoo type of class, where maybe it's the set of learnable systems, the set of learnable natural systems, LNS?
This is a Demis Hassabis new class of systems that could be actually learnable by classical systems in this kind of way, natural systems that can be modeled efficiently.
Yeah. I mean, I've always been fascinated by the P equals NP question and what is modelable by classical systems, non-quantum systems, Turing machines in effect. And that's exactly what I'm working on actually in kind of my few moments of spare time with a few colleagues about is should there be, maybe a new class of problem that is solvable by this type of neural network process and kind of mapped onto these natural systems. So, the things that exist in physics and have structure. So, I think that could be a very interesting new way of thinking about it. And it sort of fits with the way I think about physics in general, which is that I think information is primary. Information is the most fundamental unit of the universe, more fundamental than energy and matter. I think they can all be converted into each other, but I think of the universe as a kind of informational system.
So, when you think of the universe as an informational system, then the P equals NP question is a physics question.
It is a question that can help us actually solve the entirety of this whole thing going on.
Yeah, I think it's one of the most fundamental questions, actually, if you think of physics as informational. And the answer to that, I think, is going to be very enlightening.
More specific to the P and NP question, this again, some of the stuff we're saying is kind of crazy right now, just like the Christian Anfinsen Nobel Prize speech controversial thing that he said sounded crazy, and then you went and got a Nobel Prize for this with John Jumper, solved the problem. So, let me just stick to the P equals NP. Do you think there's something in this thing we're talking about that could be shown if you can do something like polynomial time or constant time compute ahead of time, and construct this gigantic model, then you can solve some of these extremely difficult problems in a theoretical computer science kind of way?
Yeah, I think that there are actually a huge class of problems that could be couched in this way, the way we did AlphaGo and the way we did AlphaFold, where you model what the dynamics of the system is, the properties of that system, the environment that you're trying to understand, and then that makes the search for the solution or the prediction of the next step efficient, basically polynomial time, so tractable by a classical system, which a neural network is. It runs on normal computers, classical computers, Turing machines, in effect. And I think it's one of the most interesting questions there is is how far can that paradigm go? I think we've proven and the AI community in general that classical systems, Turing machines, can go a lot further than we previously thought. They can do things like model the structures of proteins and play Go to better than world champion level. And a lot of people would have thought maybe 10, 20 years ago that was decades away or maybe you would need some sort of quantum systems to be able to do things like protein folding. And so, I think we haven't even scratched the surface yet of what classical systems, so-called, could do. And of course, AGI being built on a neural network system on top of a classical computer would be the ultimate expression of that. And I think the limit of what the bounds of that kind of system can do, it's a very interesting question and directly speaks to the P equals NP question.
What do you think, again, hypothetical might be outside of this? Maybe emergent phenomena? Like if you look at cellular automata, some of the extremely simple systems and then some complexity emerges. Maybe that would be outside or even would you guess even that might be amenable to efficient modeling by a classical machine?
Yeah, I think those systems would be right on the boundary. I think most emergent systems, cellular automata, things like that, could be modelable by a classical system. You just sort of do a forward simulation of it and it probably be efficient enough. Of course, there's the question of things like chaotic systems where the initial conditions really matter and then you get to some uncorrelated end state. Those could be difficult to model. So, I think these are kind of the open questions, but when you step back and look at what we've done with the systems and the problems that we've solved, and then you look at things like VEO 3 on video generation, sort of rendering, physics and lighting and things like that, really core fundamental things in physics. It's pretty interesting. I think it's telling us something quite fundamental about how the universe is structured, in my opinion. So, in a way, that's what I want to build AGI for is to help us as scientists answer these questions like P equals NP.
Yeah, I think it will maybe continuously surprise us about what is modelable by classical computers. I mean, AlphaFold 3 on the interaction side is surprising that you can make any kind of progress on that direction. Alpha genome is surprising that you can map the genetic code to the function. Kind of playing with the emergent phenomena, you think there's so many combinatorial options and then here you go. You can find the kernel that is efficiently modelable.
Yes, because there's some structure, there's some landscape, in the energy landscape or whatever it is that you can follow, some gradient you can follow. And of course, what neural networks are very good at is following gradients. And so, if there's one to follow and you can specify the objective function correctly, you don't have to deal with all that complexity, which I think is how we maybe have naively thought about it for decades, those problems. If you just enumerate all the possibilities, it looks totally intractable. And there's many, many problems like that. And then you think, it's like 10 to the 300 possible protein structures, it's 10 to the 170 possible Go positions. All of these are way more than atoms in the universe. So, how could one possibly find the right solution or predict the next step? And it turns out that it is possible, and of course, reality and nature does do it. Proteins do fold. So, that gives you confidence that there must be, if we understood how physics was doing that, in a sense, and then we could mimic that process, model that process, it should be possible on our classical systems is what the conjecture's about.
And of course there's non-linear dynamical systems, highly non-linear dynamical systems, everything involving fluid.
I had a recent conversation with Terence Tao who mathematically contends with a very difficult aspect of systems that have some singularities in them that break the mathematics. And it's just hard for us humans to make any kind of clean predictions about highly non-linear dynamical systems. But again, to your point, we may be very surprised what classical learning systems might be able to do about even fluid.
Yes, exactly. I mean, fluid dynamics, Navier-Stokes equations, these are traditionally thought of as very very difficult, intractable kind of problems to do on classical systems. They take enormous amounts of compute, weather prediction systems, these kind of things all involve fluid dynamics calculations. But again, if you look at something like VEO, our video generation model, it can model liquids quite well, surprisingly well, and materials, specular lighting. I love the ones where there's people who generate videos with clear liquids going through hydraulic presses and then it's being squeezed out. I used to write physics engines and graphics engines in my early days in gaming and I know it's just so painstakingly hard to build programs that can do that. And yet somehow these systems are reverse engineering from just watching YouTube videos. So, presumably what's happening is it's extracting some underlying structure around how these materials behave. So, perhaps there is some kind of lower dimensional manifold that can be learned if we actually fully understood what's going on under the hood. That's maybe true of most of reality.
Yeah, I've been continuously surprised by this aspect of VEO 3. I think a lot of people highlight different aspects including the comedic and the meaning and all that kind of stuff. And then the ultra-realistic ability to capture humans in a really nice way that's compelling and feels close to reality and then combine that with native audio. All of those are marvelous things about VEO 3, but the exactly the thing you're mentioning, which is the physics.
It's not perfect, but it's pretty damn good. And then the really interesting scientific question is what is it understanding about our world in order to be able to do that? Because if the cynical take with the diffusion models there's no way it understands anything. But it seems I mean I don't think you can generate that kind of video without understanding and then our own philosophical notion of what it means to understand is brought to the surface. To what degree do you think VEO 3 understands our world?
I think to the extent that it can predict the next frames in a coherent way, that's a form of understanding, right? Not in the anthropomorphic version of some kind of deep philosophical understanding of what's going on. I don't think these systems have that. But they have modeled enough of the dynamics that they can pretty accurately generate 8 seconds of consistent video that by eye is quite hard to distinguish. And imagine that in two or three more years time. That's the thing I'm thinking about and how incredible they would look given where we've come from one or two years ago. And so the rate of progress is incredible. A lot of people love all the stand-up comedians and it captures a lot of human dynamics very well and body language, but actually the thing I'm most impressed with and fascinated by is the physics behavior, the lighting and materials and liquids. It's pretty amazing that it can do that. And I think that shows that it has some notion of at least intuitive physics, how things are supposed to work intuitively, maybe the way that a human child would understand physics, as opposed to a PhD student being able to unpack all the equations. It's more of an intuitive physics understanding.
Well, that intuitive physics understanding, that's the base layer. That's the thing people sometimes call common sense. It really understands something. I think that really surprised a lot of people. It blows my mind that I just didn't think it would be possible to generate that level of realism without understanding.
There's this notion that you can only understand the physical world by having an embodied AI system, a robot that interacts with that world. That's the only way to construct an understanding of that world. But VEO 3 is directly challenging that, it feels like.
But VEO 3 is directly challenging that, it feels like.
And it's very interesting. Even if you would ask me 5, 10 years ago, I would have said even though I was a master in all of this, I would have said, well, you probably need to understand intuitive physics, like if I push this off the table, this glass, it will maybe shatter, and the liquid will spill out. So, we know all of these things. But I thought that there's a lot of theories in neuroscience, it's called action and perception where you need to act in the world to truly perceive it in a deep way. And there was a lot of theories about you need embodied intelligence or robotics or something or maybe at least simulated action so that you would understand things like intuitive physics. But it seems like you can understand it through passive observation, which is pretty surprising to me. And I think hints at something underlying about the nature of reality, beyond just the cool videos it generates. And of course, next stages is maybe even making those videos interactive, so one can actually step into them and move around them, which would be really mind-blowing, especially given my games background. And then I think we're starting to get towards what I would call a world model, a model of how the world works, the mechanics of the world, the physics of the world, and the things in that world. And of course that's what you would need for a true AGI system.
I have to talk to you about video games. So, you were being a bit trolly.
I think you're having more and more fun on Twitter on X, which is great to see. So, guy named Jimmy Apples tweeted, "Let me play a video game of my VEO 3 videos already. Google cooked so good, playable world models when?" spelled w e n. And then you quote tweeted that with, "Now, wouldn't that be something?" So, how hard is it to build game worlds with AI? Maybe can you look out into the future of video games, 5, 10 years out? What do you think that looks like?
Well, games were my first love really, and doing AI for games was the first thing I did professionally in my teenage years, and was the first major AI systems that I built. And I always want to scratch that itch one day and come back to that. I think I sort of dream about what I would have done back in the '90s if I'd had access to the kind of AI systems we have today, and I think you could build absolutely mind-blowing games. And I think the next stage is I always used to love making all the games I've made are open world games. So, they're games where there's a simulation, and then there's AI characters, and then the player interacts with that simulation, and the simulation adapts to the way the player plays. And I always thought they were the coolest games because games like Theme Park that I worked on where everybody's game experience would be unique to them. Because you're kind of co-creating the game. We set up the parameters, we set up initial conditions, and then you as the player are immersed in it, and then you are co-creating it with the simulation. But of course, it's very hard to program open world games. You've got to be able to create content whichever direction the player goes in, and you want it to be compelling no matter what the player chooses. And so, it was always quite difficult to build things like cellular automata type of those kind of classical systems which created some emergent behavior. But they're always a little bit fragile, a little bit limited. Now we're maybe on the cusp in the next few years, 5-10 years, of having AI systems that can truly create around your imagination, can dynamically change the story, and storytelling the narrative around and make it dramatic no matter what you end up choosing. So, it's like the ultimate choose your own adventure sort of game. And I think maybe we're within reach if you think of an interactive version of VEO, and then wind that forward 5 to 10 years, and imagine how good it's going to be.
Yeah, so you said a lot of super interesting stuff there. So, one, the open world built into that is a deep personalization, the way you've described it. So, it's not just that it's open world, but you can open any door and there'll be something there. It's that the choice of which door you open in an unconstrained way defines the world you see. So, some games tried to do that, they give you choice, but it's really just an illusion of choice because you're only like like Stanley Parable is a game I played. It's really there's a couple of doors and it really just takes you down a narrative. Stanley Parable is a great video game, I recommend people play, that kind of in a meta way mocks the illusion of choice and there's philosophical notions of free will and so on. But I do like one of my favorite games of Elder Scrolls is Daggerfall, I believe, that they really played with random generation of the dungeons. If you can step in and they give you this feeling of an open world. And there, you mentioned interactivity, you don't need to interact that much. You just when you open the door, whatever you see is randomly generated for you. And that's already an incredible experience because you might be the only person to ever see that.
Yeah. Exactly. And so, but what you'd like is a little bit better than just sort of a random generation. You'd like better than a simple A/B hard-coded choice, which is not really open world. As you say, it's just giving you the illusion of choice. What you want to be able to do is potentially anything in that game environment. And I think the only way you can do that is to have generated systems that will generate that on the fly. Of course, you can't create infinite amounts of game assets. It's expensive enough already how AAA games are made today. And that was obvious to us back in the '90s when I was working on all these games. I think maybe Black & White was the game that I worked on early stages of that had the still probably the best AI learning AI in it. It was an early reinforcement learning system that you were looking after this mythical creature and growing it and nurturing it. And depending how you treated it, it would treat the villagers in that world in the same way. So, if you were mean to it, it would be mean. If you were good, it would be protective. So, it was really a reflection of the way you played it. So, actually all of the work I've been doing on simulations and AI through the medium of games at the beginning of my career. And really the whole of what I do today is still a follow-on from those early more hard-coded ways of doing the AI to now fully general learning systems that are trying to achieve the same thing.
Yeah, it's been interesting, hilarious, and fun to watch you and Elon obviously itching to create games because you're both gamers. And one of the sad aspects of your incredible success in so many domains of science, like serious adult stuff, that you might not have time to really create a game. You might end up creating the tooling that others would create the game and you have to watch others create the thing you've always dreamed of. Do you think it's possible you can somehow in your extremely busy schedule actually find time to create something like Black & White, an actual video game where you could make the childhood dream come become?
Well, there's two ways to think about that. Maybe with vibe coding as it gets better, there's a possibility that I could do that actually in my spare time. So, I'm quite excited about that as that would be my project if I got the time to do some vibe coding. I'm actually itching to do that. And then the other thing is, maybe it's a sabbatical after AGI has been safely stewarded into the world and delivered into the world. And then working on my physics theory as we talked about at the beginning, those would be my two post-AGI projects, let's call it that way.
I would love to see which post-AGI would you choose, solving the problem that some of the smartest people in human history contended with, so P equals NP, or creating a cool video game?
Yeah. Well, they might be related because it would be an open world simulated game as realistic as possible. So, what is the universe? That's speaking to the same question. P equals NP, I think all these things are related, at least in my mind.
I mean, in a really serious way, so video games sometimes are looked down upon. It's just this fun side activity. But especially as AI does more and more of the difficult boring tasks, something we in the modern world called work. Video games is the thing in which we may find meaning, in which we may find what to do with our time. You could create incredibly rich meaningful experiences. Like that's what human life is. And then in video games you can create more sophisticated, more diverse ways of living.
Yeah. I think so. I mean, those of us who love games and I still do, it's almost can let your imagination run wild. I used to love games and working on games so much because it's the fusion, especially in the '90s and early 2000s, the golden era of the games industry. And it was all being discovered, new genres were being discovered. We weren't just making games, we felt we were creating a new entertainment medium that never existed before. Especially with these open world games and simulation games where you as the player were co-creating the story. There's no other entertainment media where you do that, where the audience actually co-creates the story. And of course now with multiplayer games as well it can be a very social activity and can explore all kinds of interesting worlds in that. But on the other hand, it's very important to also enjoy and experience the physical world. But then the question is, we're going to have to confront the question again of what is the fundamental nature of reality. What is going to be the difference between these increasingly realistic simulations and multiplayer ones and what we do in the real world.
Yeah, there's clearly a huge amount of value to experiencing the real world nature. There's also a huge amount of value in experiencing other humans directly in person, the way we're sitting here today. But, we need to really scientifically, rigorously answer the question why. And which aspect of that can be mapped into the virtual world?
And it's not enough to say, "Yeah, you should go touch grass and hang out in nature." It's like, "Why exactly is that valuable?"
Yes. And I guess that's maybe the thing that's been haunting me, obsessing me from the beginning of my career. If you think about all the different things I've done, they're all related in that way. This simulation, nature of reality, and what is the bounds of what can be modeled.
Sorry for the ridiculous question, but so far, what is the greatest video game of all time? What's up there? What makes?
Favorite one of all time is Civilization. I have to say. Civilization 1 and Civilization 2, my favorite games of all time.
I can only assume you've avoided the most recent one because it would probably you would that would be your sabbatical. You would disappear.
Yes, exactly. They take a lot of time, these Civilization games. So, I've got to be careful with them.
Fun question. You and Elon seem to be somehow solid gamers. Is there a connection between being great at gaming and being great leaders of AI companies?
I don't know. It's an interesting one. We both love games and it's interesting he wrote games as well to start off with. It's probably especially in the era I grew up in where home computers just became a thing in the late '80s and '90s, especially in the UK. I had a Spectrum and then a Commodore Amiga 500, which was my favorite computer ever. And that's where I learned all my programming. And of course, it's a very fun thing to program is to program games. So, I think it's a great way to learn programming. Probably still is. And then of course, I immediately took it in directions of AI and simulations, which I was able to express my interest in games and my wider scientific interests all together. And then the final thing I think that's great about games is it fuses artistic design, art with the most cutting-edge programming. So again in the '90s, all of the most interesting technical advances were happening in gaming, whether that was AI, graphics, physics engines, hardware, even GPUs of course were designed for gaming originally. So, everything that was pushing computing forward in the '90s was due to gaming. So interestingly, that was where the forefront of research was going on, and it was this incredible fusion with art, graphics, but also music and just the whole new media of storytelling. And I love that. For me, it's this multi-disciplinary kind of effort that I've enjoyed my whole life.
I have to ask you. I almost forgot about one of the many and I would say one of the most incredible things recently that somehow didn't yet get enough attention is AlphaGo Evolve. We talked about evolution a little bit, but it's the Google DeepMind system that evolves algorithms.
Are these kinds of evolution-like techniques promising as a component of future super intelligence systems? So if people don't know, it's kind of LLM-guided evolution search. So the evolutionary algorithms are doing the search and LLMs are telling you where.
Yes, exactly. So LLMs are kind of proposing some possible solutions and then you use evolutionary computing on top to find some novel part of the search space. So actually I think it's an example of very promising directions where you combine LLMs or foundation models with other computational techniques. Evolutionary methods is one, but you could also imagine Monte Carlo tree search. Basically, many types of search algorithms or reasoning algorithms on top of or using the foundation models.
models as a basis. So, I actually think there's quite a lot of interesting things to be discovered probably with these sort of hybrid systems, let's call them.
But, not to romanticize evolution, I'm only human, but you think there's some value in whatever that mechanism is? Because we already talked about natural systems. Do you think where there's a lot of low-hanging fruit of us understanding, being able to model, being able to simulate evolution and then using that whatever we understand about that nature-inspired mechanism to then do search better and better and better.
Yes. So, if you think about again breaking down the sort of systems we've built to their really fundamental core, you've got like the model of the underlying dynamics of the system. And then if you want to discover something new, something novel that hasn't been seen before, then you need some kind of search process on top to take you to a novel region of the search space. And you can do that in a number of ways. Evolutionary computing is one. With AlphaGo, we just used Monte Carlo tree search, right? And that's what found move 37, the new kind of never seen before strategy in Go. And so, that's how you can go beyond potentially what is already known. So, the model can model everything that you currently know about, right? All the data that you currently have, but then how do you go beyond that? So, that starts to speak about the ideas of creativity. How can these systems create something new, find discover something new? Obviously, this is super relevant for scientific discovery or pushing science and medicine forward, which we want to do with these systems. And you can actually bolt on some fairly simple search systems on top of these models and get you into a new region of space. Of course, you also have to make sure that you're not searching that space totally randomly, or it would be too big. So, you have to have some objective function that you're trying to optimize and hill climb towards and that guides that search.
But, there's some mechanism of evolution that are interesting. Maybe in the space of programs, but then the space of programs is an extremely important space because you can probably generalize the everything, you know, that. But, you know, for example, mutation. This is not just Monte Carlo tree search where it's like a search. You could every once in a while
combine things out there like sub like a components of a thing.
So, then, you know, what evolution is really good at is not just the natural selection. It's combining things and building increasingly complex hierarchical systems. So, that component is super interesting, especially like with Alpha Evolve in the space of
Yeah, exactly. So, there's a you can get a bit of an extra property out of evolutionary systems, which is some new emergent capability may come about. Right? Of course, like it happened with life. Interestingly, with naive, sort of traditional evolution competing methods without LLMs and the modern AI, the problem with them they there was a they were very well studied in the '90s and early 2000s and some promising results, but the problem was they could never work out how to evolve new properties. New emergent properties. You always had a sort of subset of the properties that you put into the system, but maybe if we combine them with these foundation models, perhaps we can overcome that limitation. Obviously, natural evolution clearly did because it did evolve new capabilities. Right? So, bacteria to where we are now. So, clearly that it must be possible with evolutionary systems to generate new patterns, going back to the first thing we talked about and new capabilities and emergent properties. And maybe we're on the cusp of discovering how to do that.
Yeah, listen, AlphaFold is one of the coolest things I've ever seen. I've On my desk at home, you know, most of my time is spent behind a computer just programming. And next to the three screens is a skull of a Tiktaalik, which is one of the early organisms that crawled out of the water onto land. And I just kind of watch that little guy. [laughter] It's like you whatever the competition mechanism of evolution is is quite incredible. Yeah, it's truly truly incredible. Now, whether that's exactly the thing we need to do to do our search, but never never dismiss the power of nature with what it did here.
Yeah. And it's amazing. With a relatively simple algorithm, right, effectively, and it can generate all of this immense complexity. It emerges. Obviously running over, you know, 4 billion years of time, but it's you know, you can think about that as again a process that ran over the physics substrate of the universe for a long amount of computational time. But then it generated all this incredible rich diversity.
So, so many questions I want to ask. So, one, you do have a dream. One of the natural systems you want to try to model is a cell. That's a beautiful dream. I could ask about that. I also just for that purpose on the AI scientist front just broadly. So, there's an essay from Daniel Kokotajlo, Scott Alexander, and others that outline steps along the way to get to ASI. And has a lot of interesting ideas in it. One of which is including a superhuman coder and a superhuman AI researcher. And in that there's a term of research taste. That's really interesting. So, in everything you've seen, do you think it's possible for AI systems to have research taste? To help you in the way that AI co-scientist does to help steer human brilliant scientists and then potentially by itself to figure out what are the directions where you want to generate truly novel ideas? Because that seems to be like a really important component of how to do great science.
Yeah, I think that's going to be one of the hardest things to mimic or model is this idea of taste or judgment. I think that's what separates the great scientists from the good scientists. Like all professional scientists are good technically, right? Otherwise, they wouldn't have made it that far in academia and things like that. But then, do you have the taste to sort of sniff out what the right direction is, what the right experiment is, what the right question is? So, picking the right question is the hardest part of science. And making the right hypothesis. And that's what today's systems definitely can't do. So, you know, I often say it's harder to come up with a conjecture, a really good conjecture, than it is to solve it. So, we may have systems soon that can solve pretty hard conjectures. You know, I am in math olympiad problems. Well, we you know, AlphaProof last year, our system got you know, silver medal in that. Really hard problems. Maybe eventually we'll be able to solve a millennium prize kind of problem. But could a system have come up with a conjecture worthy of study that someone like Terence Tao would have gone, you know what? That's a really deep question about the nature of maths or the nature of numbers or the nature of physics. And that is far harder type of creativity. And we don't really know today's systems clearly can't do that. And we're not quite sure what that mechanism would be. This kind of leap of imagination. Like Einstein had when he came up with special relativity and then general relativity with the knowledge he had at the time.
As for conjecture, the you want to come up with a thing that's interesting is amenable to proof.
So like it's easy to come up with a thing that's extremely difficult. Yeah. It's easy to come up with a thing that's extremely easy, but that at that very edge
That sweet spot, right? Of basically advancing the science and splitting the hypothesis space into two ideally, right? Where if it's true or not true, you've learned something really useful. And that's hard. And making something that's also falsifiable and within sort of the technologies that you currently have available. So it's a very creative process actually, highly creative process that I think just a kind of naive search on top of a model won't be enough for that.
Okay, the idea of splitting the hypothesis space is super interesting. So I've heard you say that there's basically no failure or failure is extremely valuable if it's done if you construct the questions right, if you construct the experiments right, if you design them right, that failure or success are both useful. So perhaps because it splits the hypothesis space into two, it's like a binary search.
Yes. That's right. So when you do like, you know, real blue sky research, there's no such thing as failure really as long as you're picking experiments and hypotheses that meaningfully split the hypothesis space. So, you know, and you learn something you can learn something kind of equally valuable from an experiment that doesn't work. That should tell you if you've designed the experiment well and your hypotheses are interesting, it should tell you a lot about where to go next. And then you're effectively doing a search process and using that information in very helpful ways.
So to go to your dream [snorts] of modeling a cell, what are the big challenges that lay ahead for us to make that happen? We should maybe highlight that AlphaFold I mean, there's just so many leaps.
So, AlphaFold solved, if it's fair to say, protein folding, and there's so many incredible things we could talk about there, including the open sourcing. The everything you've released. AlphaFold 3 is doing protein RNA DNA interactions,
which is super complicated and fascinating that's amenable to modeling. AlphaGenome predicts how small genetic changes, like if we think about single mutations, how they link to actual function. So, those are it seems like it's creeping along to a sophisticated to much more complicated things like a cell. But, a cell has a lot of really complicated components.
Yeah. So, what I've tried to do throughout my career is I have these really grand dreams, and then I try to, as you've noticed, and then I try to break them down, you know, it's easy to have a kind of a crazy ambitious dream, but the trick is how do you break it down into manageable, achievable interim steps that are meaningful and useful in their own right. And so, virtual cell, which is what I call the project of modeling a cell, I've had this idea, you know, of wanting to do that for maybe more like 25 years. And I used to talk with Paul Nurse, who is a bit of a mentor of mine in biology. He runs the you know, founded the Crick Institute and won the Nobel Prize in 2001. It is we've been talking about it since, you know, before the year you know, in the '90s. And I used to come back to every 5 years. It's like, what would you need to model of the full internals of a cell so that you could do experiments on the virtual cell, and what those experiments, you know, in silico, and those predictions would be useful for you to save you a lot of time in the wet lab, right? That would be the dream. Maybe you could 100x speed up experiments by doing most of it in silico. The search in silico and then you do the validation step in the wet lab. That would be the dream. And so, but maybe now finally so I was trying to build these components AlphaFold being one that would allow you eventually to model the full interaction a full simulation of a cell. And I'd probably start with the yeast cell and partly that's what Paul Nurse studied because the yeast cell is like a full organism that's a single cell, right? So, it's the kind of simplest single cell organism. And so, it's not just a cell it's a full organism. And yeast is very well understood. And so, that would be a good candidate for a kind of full simulated model. Now, AlphaFold is the solution to the kind of static picture of what does a protein look 3D structure of protein look like? A static picture of it, but we know that biology all the interesting things happen with the dynamics, the interactions. And that's what AlphaFold 3 is is the first step towards is modeling those interactions. So, first of all pairwise, you know, proteins with proteins, proteins with RNA and DNA. But then the next step after that would be modeling maybe a whole pathway, maybe like the TOR pathway that's involved in cancer or something like this. And then eventually you might be able to model a whole cell.
Also, there's another complexity here that stuff in a cell happens at different time scales. Is that tricky? It's like the you know, protein folding is you know, super fast.
I don't know all the biological mechanisms.
But some of them take a long time. And so, is that that's a level So, the levels of interaction has a different temporal scale that you have to be able to model.
So, that would be hard. So, you'd probably need several simulated systems that can interact at these different temporal dynamics. Or at least maybe it's like a hierarchical system. So, you can jump up and down the different temporal stages.
So, can you avoid I mean one of the challenges here is not avoid simulating. For example, the quantum mechanical aspects of any of this, right? You want to not over model. You could skip ahead to just model the really high-level things that get you a really good estimate of what's going to happen.
You go to make a decision when you're modeling any natural system, what is the cutoff level of the granularity that you're going to model it to that then captures the dynamics that you're interested in. So, probably for a cell, I would hope that would be the protein level. And that one wouldn't have to go down to the atomic level. So, you know, of course, that's where AlphaFold kicks in. So, that would be kind of the basis and then you'd build these higher-level simulations that take those as building blocks and then you get the emergent behavior.
Apologies for the pothead questions ahead of time, but will do you think we'll be able to simulate a model the origin of life? So, being able to simulate the first from non-living organisms, the birth of a living organism.
I think that's one of the of course, one of the deepest and most fascinating questions. I love that area of biology, you know, there's people like there's a great book by Nick Lane, one of the top experts in this area called the 10 great inventions of evolution. I think it's fantastic and it also speaks to what the great filters might be, but you know, prior or are they ahead of us? I think they're most likely in the past if you read that book of how unlikely to go, you know, have any life at all and then single cell to multi-cell seems an unbelievably big jump that took like a billion years, I think, on Earth to do, right? So, shows you how hard it was.
Bacteria were super happy for a very long time.
Time before they captured mitochondria somehow, right? I don't see why not why AI couldn't help with that. Some kind of simulation again, it's again, it's a bit of a search process through a combinatorial space. Here's like all the chemical soup that you start with, the primordial soup that you know, maybe was on Earth near these hot vents. Here's some initial conditions. Can you generate something that looks like a cell? So perhaps that would be a next stage after the virtual cell project is well, how could you actually something like that emerge from the chemical soup?
Well, I would love it if there was a move 37 for the origin of life.
I think that's one of the sort of great mysteries. I think ultimately what we will figure out is there a continuum. There's no such thing as a line between non-living and living. But if we can make that rigorous.
That the very thing from the big bang to today has been the same process. If we can break down that wall that we've constructed in our minds of the actual origin of from non-living to living and it's not a line that it's a continuum that connects physics and chemistry and biology.
I mean, this is my whole reason why I worked on AI and I'm my whole life because I think it can be the ultimate tool to help us answer these kind of questions. And I don't really understand why the average person doesn't think like worry about this stuff more. Like how [laughter] how can we not have a good definition of life and not living and non-living and [laughter] the nature of time and let alone consciousness and gravity and all these things. It's just and quantum mechanics weirdness. It's just to me it's I've always had this sort of screaming at me in my face. And it's getting louder. You know, it's like how what is going on here? In the deepest sense. Like the nature of reality which has to be the ultimate question.
That would answer all of these things. It's sort of crazy if you think about it. We can stare each other and all these living things all the time. We can inspect it in microscopes and take it apart almost down to the atomic level and yet we still can't answer that clearly in a way that question of how do you define living?
Yeah. It's kind of amazing.
Yeah, living you can kind of talk your way out of thinking about but like consciousness like we have this very obviously subjective conscious experience like we're at the center of our own world and it feels like something and then how are you not screaming
at the mystery but all whatever. I mean but really humans have been contending with the mystery of the world around them for a long long There's a lot of mysteries like what's up with the sun and the rain
like what's that about and then like last year we had a lot of rain and this year we don't have rain like what did we do wrong? Humans have been asking that question for a long time.
Exactly. So we're quite I guess we've developed a lot of mechanisms to cope with these deep mysteries that we can't fully we can see but we can't fully understand and we have to just get on with daily life and we keep ourselves busy right? In a way do we keep ourselves distracted?
I mean weather is one of the most important questions of human history. We still it's that's the go-to small talk direction of the weather.
And then it's which is you know, famously is an extremely difficult system to model and even that system the Google DeepMind has made progress on.
Yes, we've created the best weather prediction systems in the world and they're better than traditional fluid dynamics sort of systems that usually calculated on massive supercomputers takes days to calculate it we've managed to model a lot of the weather dynamics with neural network systems with our weather next system and again it's interesting that those kinds of dynamics can be modeled even though they're very complicated almost bordering on chaotic systems in some cases a lot of the interesting aspects of that can be modeled by these neural network systems including very recently we had you know, cyclone prediction of where, you know, parts of hurricanes might go. Of course, super useful, super important for the world. And it's super important to do that very timely and very quickly and as well as accurately. And I think it's very promising direction again of, you know, simulating and so they can run forward predictions and simulations of very complicated real-world systems.
As you mentioned that I've got a chance in Texas to meet a community of folks called the storm chasers.
And what's really incredible about them, I need to talk to them more, is they're extremely tech-savvy. Because what they have to do is they have to use models to predict where the storm is.
So they're it's this [laughter] it's this beautiful mix of like crazy enough to like go into the eye of the storm and like in order to protect your life and predict where the extreme events are going to be, they have to have increasingly sophisticated models of weather.
Yeah, it's a beautiful balance of like being in it as living organisms and the cutting edge of science. So, they actually might be using DeepMind system. So, that's
Yeah, they I would hopefully they are and I'd love to join them on one of those chases. [laughter] They look amazing, right? To actually experience it one time.
Exactly. And then also to experience the correct prediction where something will come and how it's going to evolve. It's incredible.
You've estimated that we'll have AGI by 2030. So, there's interesting questions around that. How will we actually know that we got there? And what may be the move, quote, move 37 of AGI.
My estimate is sort of 50% chance by in the next 5 years. So, you know, by 2030, let's say. And so I think there's a good chance that that could happen. Part of it is what is your definition of AGI? Of course, people are arguing about that now and mine's quite a high bar and always has been of like can we match the cognitive functions that the brain has, Right, so we know our brains are pretty much general Turing machines, approximate, and of course we created incredible modern civilization with our minds. So, that should also speak to how general the brain is. And for us to know we have a true AGI, we would have to like make sure that it has all those capabilities. It isn't kind of a jagged intelligence where some things it's really good at, like today's systems, but other things it's really flawed at. And that's what we currently have with today's systems. They're not consistent. So, you'd want that consistency of intelligence across the board. And then we have some missing, I think, capabilities. Like sort of the true invention capabilities and creativity that we were talking about earlier. So, you'd want to see those. How you test that? I think you just test it. One way to do it would be a kind of brute force test of tens of thousands of cognitive tasks that, you know, we know that humans can do. And maybe also make the system available to a few hundred of the world's top experts, the Terrence Taos of each subject area, and see if they can find, you know, give them a month or two and see if they can find an obvious flaw in the system. And if they can't, then I think you're pretty confident you have a fully general system.
Maybe to push back a little bit. It seems like humans are really incredible as the intelligence improves across all domains to take it for granted. Like you mentioned Terrence Tao, these brilliant experts, they might quickly, in a span of weeks, take for granted all the incredible things it can do and then focus in well, 'Haha, right there.' You know, I consider myself a first of all human.
I identify as human. Is it I you know, some people listen to me talk and they're like, 'That guy's not good at talking. The stuttering, the you know' [laughter] So like even humans have obvious across domains limits. Even just outside of mathematics and physics and so on. I wonder if it will take something like a move 37. So on the positive side versus like a barrage of 10,000 cognitive tasks where it would be one or two where it's like
Exactly. So I think there's the sort of blanket testing to just make sure you've got the consistency. But I think there are the sort of lighthouse moments like the move 37 that I would be looking for. So one would be inventing a new conjecture or a new hypothesis about physics like Einstein did. So maybe you could even run the back test of that very rigorously. Like have a cutoff of knowledge cutoff of 1900 and then give the system everything that was written up to 1900 and then see if it could come up with special relativity and general relativity, right? Like Einstein did. That would be an interesting test. Another one would be can it invent a game like Go? Not just come up with move 37 a new strategy, but can it invent a game that's as deep as aesthetically beautiful as elegant as Go? And those are the sorts of things I would be looking out for. And probably a system being able to do several of those things, right? For it to be very general. Not just one domain. And so I think that would be the signs at least that I would be looking for that we've got a system that's AGI level. And then maybe to fill that out you would also check their consistency. You know, make sure there's no holes in that system either.
Yeah, something like a new conjecture or scientific discovery. That would be a cool feeling.
Yeah. That would be amazing. So it's not just helping us do that, but actually coming up with something brand new.
And you would be in the room for that. And so it would be like probably two or three months before announcing it.
And you would just be sitting there trying not to tweet.
