A Conversation with Demis Hassabis, Co-Founder and CEO of Google DeepMind

[00:00:00] Speaker 1: And it's such a pleasure to see everybody here for the conversation with Demis Hassabis. We're especially honored today to have President John Levin leading this fireside chat. One of the things that makes Stanford University so distinctive is that some of the most important ideas emerge, not within a single school or discipline, but at the intersections of schools and units. That spirit of cross-university collaboration is particularly important right now as advances in AI begin to reshape nearly every domain of society. And nowhere is this more consequential than in medicine. I feel that personally through the GSB's close partnership with the Stanford School of Medicine, which is undertaking an extraordinary effort to reimagine cancer innovation and care by bringing together social scientists, scientists, clinicians, engineers, and innovators to transform the patient journey from prevention through survivorship. That vision is ambitious, and it will require the full capabilities of a great university working together. Stanford's strength lies not just in excellence within fields, but in our ability to connect fields. We bring AI researchers together with physicians, organizational leaders together with scientists, and entrepreneurs together with people deeply committed to human well-being. And that is one reason that today's conversation feels so important. Demis Hassabis is an artificial intelligence researcher, an entrepreneur, and a Nobel laureate whose work sits exactly at these intersections. He's the co-founder and CEO of Google DeepMind, one of the world's leading AI research companies. This was founded as DeepMind in 2010 and acquired by Google in 2014, and the company is now central to Google's AI efforts and has produced some of the defining breakthroughs in the field. Some of these breakthroughs include AlphaGo, the first program to defeat a world champion at the game of Go, and AlphaFold, which solved the 50-year grand challenge of protein structure prediction by accurately predicting the three-dimensional shapes of proteins. This was a breakthrough with enormous implications for disease understanding and drug discovery. For this work, Demis, alongside John Jumper and David Baker, was awarded the 2024 Nobel Prize in Chemistry. He's also a fellow of the Royal Society and the Royal Academy of Engineering. And in 2024, he was knighted for services to artificial intelligence. He has also been named the Time 100 list of the world's most influential people multiple times, including in 2017 and 2025. But what makes this moment especially compelling at Stanford is that the conversation around AI here has never been only about capability. It has also been about human flourishing. Several years ago, professors Fei-Fei Li and Jennifer Auker began teaching a Stanford course on AI for human flourishing built around a profound set of questions. What does it mean to be human, what does flourishing look like, and when does technology help advance those goals versus undermine them? One insight from this work has stayed really deeply with me, and this is that some forms of friction are actually load-bearing. The struggle to find the right word, the discomfort of difficult conversations, the challenge of learning something new, are not inefficiencies to eliminate. They are instead the very experiences through which growth, agency, resilience, and meaning emerge. And that, too, is why our discussion today matters so much. At Stanford, advances in AI are not abstract. They are already reshaping how we think about discovery, diagnosis, leadership, learning, and human potential itself. Advances in AI are also forcing us to grapple with larger questions about judgment, ethics, institutions, and what kinds of lives we ultimately want technology to help us build. So thank you all for being here, and please join me in welcoming President John Levin and Demis Hassibis to the stage. Thank you. [00:05:11] John Levin: Demis, it's great to have you here at Stanford. [00:05:13] Demis Hassabis: It's fantastic to be here. Thanks, everyone, for coming. [00:05:15] John Levin: Really appreciate you doing this. So we're going to ask you some questions. We'll have some questions from students, and we're looking forward to hearing your thoughts. Maybe we would, you've been chronicled a lot recently, a movie, a book. So many people have heard about your trajectory. It is quite remarkable. Chess prodigy, video game developer, scientist, tech entrepreneur and leader, Nobel laureate. That's just the first half of your career. So if you were going to try to draw a through line through all those different things that you've done, what would it be? [00:06:03] Demis Hassabis: Well, I think there's several through lines, actually, with what seems maybe somewhat unconnected subjects. First of all, I've always really enjoyed working at the intersection of creativity and technology, and very broadly construed. So actually, the games industry, the video games industry, which is the first, my first very early part of my career, in the 90s, was one of the most creative spaces in any industry that was using cutting edge technology with art and design to sort of create an entirely new entertainment medium. So that was really an amazing time. In fact, some of the most fun times I've had in my career was early in the 90s. The chess and the neuroscience I did, all of those things I've tried to, I had from this idea of working on AI and AGI being the most important thing one could, and most interesting one thing one could spend your career working on, also from a very early age. So as a teenager, probably I read too much science fiction, reading things like Gödel, Escher, Bach, these types of books, and biographies of some of my scientific heroes, you know, Turing and Feynman and so on. So all of these were serving to inspire me to try and understand the world around us in a really deep way. And then building AI was my expression of that mission, to try and build the ultimate tool for science. And I've tried to, because life's short, I've tried to reuse and repurpose every experience I've had in service of that bigger North Star mission that I've had, you know, for more than 30 years. So, you know, my chess training is the way that I think about business and organizing things and planning and how I think I've been able to break down very ambitious plans into smaller, more manageable steps. That all comes from kind of chess thinking, I would say. And then using games, first of all, building games, learning about engineering projects at scale, running companies, startups, and then fusing this creativity with engineering. Actually, it's what we do today with AI. It's an engineering science. So you're fusing creative work, scientific work, with very hardcore, cutting-edge engineering. So that all served together. And then finally, on games, as everyone knows, we used games in the early days of DeepMind as the perfect proving ground for testing out algorithmic ideas, probably most famously with AlphaGo, which I think, you know, we've just had the 10-year anniversary of. And it was really, looking back now, maybe the start of the modern AI era. [00:08:38] John Levin: When you went into AI professionally in 2010 or so, you started DeepMind, you had this very ambitious vision. You were going to solve intelligence and then solve everything else. Yeah. How's it going? Well, let me expand a little bit. What has gone according to plan and sort of what has been off plan? [00:09:04] Demis Hassabis: Well, the broad arcs of it have gone, I mean, unbelievably well. Perhaps, you know, when we started DeepMind in 2010, that's, can you imagine, we used to go to, try to go to VCs in the UK, in which there weren't very many. And with that as the business plan, it was literally step one, solve intelligence, step two, use it to solve everything else. And people were quite confused. But we really meant it. And actually, we may go back to exactly using that mission statement, because by solve intelligence, we meant build AGI. Ideally, also understand the nature of intelligence on the way to building AGI and perhaps using AGI to help us understand our own brains. And minds better, you know, things like the nature of consciousness, what is creativity, dreaming, all of these deep mysteries of the mind. And one of the reasons I studied neuroscience was try to learn from what we understood about the brain as inspiration for algorithmic ideas. And sort of, so step one was to try and build AGI. And then we always had in mind what sort of happened, which is that, of course, it's a general purpose technology, maybe the general purpose technology. And if it was built in the right way, so it was a learning system that was very general, what would be the limit of what it could be applied to? It's not, you know, it could be applied to almost anything was the dream. And I think that's what's born out. I had specifically in mind for that step two, advancing science and medicine. So that's what I meant by using it to solve everything else. I meant the big questions in science, all of them. I wanted to, you know, I was fascinated by all of them, the nature of time, the nature of reality. Maybe that's the most fundamental one. And I loved physics when I was at school. That was my favorite subject. And I think when you're interested in the big questions, you end up doing physics, probably. But the reason I decided that there were too many interesting big questions. So how was one to, you know, try and tackle all of that in a lifetime? And that meant, in my view, building new tools and to aid us at the best scientists, the best experts, to make much faster progress in the fields that they were tackling and the big questions and important questions that they were tackling. And then, of course, AI in itself is also a fascinating artifact in itself, scientific sort of object, one could call, worthy of study itself. It's almost a new field. So this is sort of, to me, like the most fascinating and most important thing to spend one's life on. And I would have been doing it, even if it hadn't worked out, I would have found some way to be doing this, you know, in academia or wherever. This is what I would always plan to spend my life working on. And all those things I did earlier were different expressions, gathering the experience and, I suppose, the knowledge to be able to attempt something like DeepMind in 2010 when we were sort of, I felt, we felt that we were ready to make fast progress. And, of course, the second part of that, use it to solve everything else, is now much broader than just science and medicine, although that's where I've tried to personally do my work in, as well as running the overall organization. But, obviously, it's going to be amazing for productivity and many other things in the world outside of just science and medicine. [00:12:29] John Levin: As you've been building these different models at DeepMind, you started with games and then you went into science. Have there been, were there particular moments where, that, I mean, you started with a lot of conviction, but I'm curious if there were sort of particular moments where you sort of saw this is actually going to work. Yeah. With the alpha go and playing move third. [00:12:55] Speaker 4: Yeah, there were many moments where I thought it wasn't going to work, put it that way. [00:12:58] Demis Hassabis: So, some of the ones I remember really well are, we started with games because they're self-contained. They obviously were designed to be, by other humans, to be challenging or fun for other humans to play. They're often, I love games, they're often microcosms of a lot of real world scenarios. You know, if you think of Go or poker or chess, I often thought as, you know, one of the courses I'd have in an MBA or business school course would be a games module to, you know, study those types of games. Diplomacy, they all have really interesting aspects of the best games of real life and you can obviously practice many times in a kind of safe scenario. That's what I think games are really useful for, and that applies to AI systems that are learning too, that, you know, they're neat environments, they're challenging, and they have clear objective functions, which is also was very important for our early days of reinforcement learning. Almost no one had used reinforcement learning for any kind of scaled up problem. It was just being used, you know, it was very kind of, it was obviously an academic discipline, but it was used mostly for toy problems like little grid worlds. Um, it wasn't clear it could scale up to anything, uh, major. Um, and so we started with, um, I would say the, the, the most, uh, famous set of, but most basic games that had become, you know, world popular, which were Atari games and, um, from the seventies. And, uh, we started with the simplest game of all, which was Pong. You know, there's just the two, the bat and the ball, just two bats and a ball. And, um, there's an inbuilt AI system, it's not really an AI system, an inbuilt system that controls your opponent and, uh, uses all the information that the game has about where the ball is and so on to move the bat around. And what we wanted to do was could you play Pong just from the pixels on the screen? So the raw data, the raw visual input and no other information, no privileged information about no access to the, the insides of the program about, you know, where the ball is or the speed of it and so on, which obviously the program knows. But we didn't give the, uh, the DQN system as it was called our Atari system, any of that information. It just got the 20,000 pixels on the screen and 20,000 pixels, I mean, it seems, uh, uh, you know, sort of trivial now, but back in 2010, that was an enormous amount of input data. No one had ever dealt with something that complex and then multiplied by all the frames that you were doing. And for about, it felt like six months, maybe it was only two months, but we couldn't win a single point at POP. So it was jerking the bat around. It was like, oh, is it ever going to even be able to control the bat? And of course, he had no notions of any of these things. And it was just losing 21-0 to the inbuilt AI. And I did think, and we had a couple of different ways of trying to attack this. And we had almost no money, you know, the runway, you know, the bare couple of million dollars of funding that we had, which wouldn't even cover an intern these days, which is good for all of you students, was our entire funding. And, you know, we were paying, we were taking, like, no salary. And it was running out, the money. And I was like, oh, well, maybe, it turns out, maybe we are still 10 years too early. Maybe we're 20 years too early. And then, magically, it got a point, and it was like, oh, maybe it was just luck. And then it started winning a lot of points, and then it started winning the games. And then it was like, okay, we have liftoff now. So now, and those of you working in machine learning will know this, if you get a foothold, you can usually hill climb your way out of that. That's been the history of AI, I would say, right? Once you have something working, there's usually a way of optimizing it more. And that's what turned out with Atari. And then, so that was our first big result, and our first nature paper was the, you know, really the first deep reinforcement learning model, certainly at scale. Combining deep learning to learn the domain and deal with the perceptual inputs and the complexity of the inputs, find the patterns in it, and then reinforcement learning built on top of that to kind of make the decisions and do the planning. And then, of course, that culminated in AlphaGo, which was the, always our aim, Dave Silva and I was head of that project. We were undergrad, friends as undergrads at Cambridge, and we were discussing it since our undergrad about, you know, we were there in the mid-90s. The Deep Blue Kasparov match happened while we were at college. Of course, I was fascinated by it both from the chess and the AI point of view, but I was more impressed with Kasparov's brain than I was with Deep Blue, because Kasparov, you know, with his incredible mind, still one of the biggest chess geniuses there's been of all time, he was able to basically compete on an equal footing with this supercomputer brute force machine next to him. But, of course, he could do all the other things with his mind, speak five languages, do his politics, drive cars, all the rest of the things humans can do. And to me, that was, you know, incredible. That's much more impressive. So there was something missing from the Deep Blue system. And obviously, those techniques, those expert system techniques, where you hand-curate the heuristics and then you use brute force search on top, which is still how a lot of traditional chess programs work today, that works for chess, but it's never worked for Go. Because Go's too esoteric a game, it hasn't got material, every piece is worth the same, it's all about patterns and intuition, even the top Go players, that's how they play it. So we realized, we sort of thought, okay, if someone could actually get to world champion level at Go, it's not just about, really, that was an aside getting to that level, it was more about the approach you would have taken would probably be a really interesting algorithmic approach, and maybe and hopefully would generalize to other domains. And that's what turned out with AlphaGo. So, and then it went beyond our wildest dreams, really, because not only did it win the match against Lisa Dole in 2016, it also created new, famously, new strategies that had never been seen before, even though we've played Go. Go is the oldest game humanity has invented, 2,000 years old, 2,000 plus years old, and been played professionally for hundreds of years. And we hadn't discovered those strategies. So I was sort of, that was sort of double whammy for me, I was waiting for that moment, that AI was able to come up with something novel, and we, you know, it's not, there's more levels of creativity than that, but beyond that, but at least it was a novel idea. And then that was, for me, was what I was waiting for, to then start using AI for science. So the moment we got back from Sol, we started the AlphaFold project. [00:19:34] John Levin: Now when you, so let's talk about the science a bit, because then you went into the protein folding problem. And again, you picked a problem where there was data and where there was a clear objective function in terms of thinking about protein folding. And it, and it, and it, and it worked that you, I mean, you actually managed to solve this longstanding problem of, of predicting protein structure. You did something very interesting when, when you, when you came up with AlphaFold, which was, it was obviously a huge science, Nobel worthy scientific breakthrough, probably also of commercial value. And you just gave it away for free. I am curious, how did you come to that decision? Did you, was it something, did you think about other ways of going about what, why, why give it away? [00:20:25] Demis Hassabis: Yeah, so we picked the, the, the protein folding problem. I, I, I had my kind of eye on that since also, you know, my undergrad days at Cambridge, just when I first came across it, I had a few biologist friends who were obsessed with the protein folding problem. And actually they, they ended up becoming structural biologists of course in their career. And, uh, one specifically I remember he was, you know, every time we were in the pub playing table football or something, he would be talking about obsessively at how this was the most important problem in biology. And more importantly, I think of it as a root note problem. Like if you could, if you could unlock that and, and, and, and find the structures of proteins that would unlock whole, you know, new avenues of research, things like drug discovery. Obviously we're trying to X, X push that, but also fundamental biology and disease understanding. So this was, um, this was, uh, a problem worth really spending a lot of attention and time on because of the downstream effects that it would have. It also felt to me, it was a fascinating problem. It felt to me like the ultimate puzzle, you know, 3d puzzle of how does, uh, uh, this sort of, you know, amino acid sequence. You can think of it as genetic sequence fold up into this 3d structure. It's amazingly interesting, uh, intricate thing. And the more I looked into proteins, the more incredible at my respect and wonder is for biology. Like these unbelievable little bio nanomachines, you know, everything on life obviously depends on proteins. And as you start looking at their structure, you start understanding their function. So this is fascinating to me as a science question. Um, and then yes, there was the clear objective is sort of like minimizing the free energy in the system. Presumably this is how physics is doing. It's why the, the, the body, you know, these proteins fold in milliseconds in your body, you know, billions of times a second. Um, so somehow physics has solved this, so it can, it can't be, there must be some, uh, topology, let's say that you could learn maybe with a deep learning system that would guide the search just like we've done with alpha go to find a great moving go a great strategy out of the, you know, more possibilities. Then there are atoms in the universe and go and protein folds are even larger search space than that, but there's some way to narrow that down in a sensible way. You learn a kind of heuristic using the deep learning models to then guide your search for that to become tractable. And it felt like a, a really, uh, analogous problem, uh, in science to what we'd solved in go, um, sort of applying some of those same, uh, approaches, those same theories to, to this domain. Um, and then the other thing was there was obviously 50 years worth of painstaking crystallography, structural biology work by many great labs and people. And, uh, uh, they are, after all of that effort, um, there was about 150,000 structures in the PDB, the main database, which isn't a lot actually. Obviously it's, it's a huge amount of effort that's gone into that, but there's 200 million proteins and 150,000 also for, for machine learning systems is a very small amount of data. So most people thought it was at least 10, 20 years away, um, before we would have enough, uh, data and the right types of algorithms to, to tackle that. But we, we felt that using every technique we knew in the end that we could make progress with that. Um, and it turned out to be the case, uh, and then, um, when we decided to, well, how would we make the maximum impact with this, it was obvious to me that we should, um, fold all the proteins so that, cause not only was alpha fold accurate, it was extremely fast. It could fold a protein in a matter of seconds and then collaborate with, in the end European bioinformatics Institute, uh, in Cambridge, uh, to, which hosts many of the biggest databases, biology databases, scientists use, uh, and just host the entire 200 million protein structures on their database and just allow it to be as simple as a kind of Google search. Um, to just find your protein structure, uh, along with the confidence intervals, the machine learning system had about which parts of the protein structure it was confident on, which is very important for biologists to know. So we put that all together, um, and it was, it was, uh, of course, it could have been very valuable. Um, I don't know how many billions of dollars or whatever. I mean, it depends how you calculate it to do that experimentally would be incalculable cost, but it would have been hugely valuable to keep proprietary. But for us, it felt, um, like we would only be able to scratch the surface of the downstream impact that putting all those structures out in the world could have, uh, on our own because, you know, there's 3 million researchers around the world that use alpha fold. Uh, pretty much every day, almost every biologist, uh, medical researcher in the world, there's no way one organization could have, could have done that. So it was obviously the right thing to do. Uh, we also had depended on public data to train the first version of alpha fold. So it only felt right to give back to that community, the structural biology community, um, this amazing resource that was, that was amplifying the resource they had basically built. And, um, so it was just, it wasn't even a question for me and, and, you know, um, it was great that also the executives at Google also, you know, love science and totally got that. Um, I don't think all companies would have made that decision. Um, so I give them a lot of QDOS on that too. That was an easy discussion. Um, and then we've tried to ourselves push that downstream with isomorphic labs, uh, an alphabet spinout that is, um, building, you can think, uh, several more type of alpha fold level breakthroughs, putting them together into a way that will accelerate, hopefully, drug discovery, you know, take it down from years to months, maybe even one day weeks, just like we did with, um, protein structures, which used to take, you know, years for a single one. And then we could do it in seconds. [00:26:02] John Levin: And that's one of the really exciting areas of the, the future with AI, I, I want to turn for a minute to something you said earlier this week. Uh, you, you, you were in the news this week because at a big Google event, you, you said, um, uh, that we're in the foothills of the singularity. [00:26:23] Demis Hassabis: Um, yes, it got quite a lot of pickup that line. [00:26:27] John Levin: It, it, it, it got a lot of, it got a lot of pickup. I, I, I, and I understand the, maybe the, maybe the, the, the Google press team might not have been so thrilled about it, but since you're out there saying that, um, what did you mean by that? [00:26:41] Demis Hassabis: Yeah. So the full thing I said to sort of close the, the, the conference with was, um, when we look back at this time, I think that, you know, maybe I'm thinking sort of 10 years from now, I think we will realize that we were standing in the foothills of the singularity. Now what I mean by that, and the way I, the reason I chose that word is that, um, so there's the technology, which is AGI we've been calling AGI this, this, this next version of really general, uh, artificial intelligence. I believe that we're only a few years away from that, maybe like 20, 30 plus or minus a year, which is astounding to think really. Um, and then the era, I think that will be such an enormous transformative technology. It's going to effectively be a new human era. And that's what I'm meaning by the singularity is that, and what many science fiction writers have written about that is that it's sort of the, it's describing the era that we will be in, um, in and around when AGI, the advent of AGI happens. So, and I think we can feel this year, I would say, even though I've been working towards this for 30 years, I think this year with the way the agents are working and tool use, um, it started to become really useful for, you know, still, early days of it, but genuinely useful in people's workflows. Um, and we can sort of see what extra things that need to be done. And all of us, the leading labs are working on that. Um, I think this is the beginnings of that, but the foothills, I still think there's a lot more work and it's just the beginnings, but, and it's not any one thing. It's, it's several different technologies, several use cases that I see, um, several things that I thought were maybe a bit further out, turned out to be now, um, that are coming together, uh, that make me feel that, uh, in aggregate. And, and, and to the extent that I wanted to say that, um, because I think, um, society needs to hear that because we don't have long to prepare for what that means. Um, and, uh, it's gonna be enormously profound, I think. Uh, and the future, in my view, is still to be written, but these next few years are gonna be very critical as to which way that will go and how we collectively, uh, want that to, to look like. [00:28:48] John Levin: If you look at surveys, uh, of how people perceive AI in, in this country in particular, uh, it's very negative right now. Uh, it's, and, uh, maybe more negative here than, than even in, in other, other countries. And, and there's probably lots of things driving that concerns about privacy or state control or the size of the tech companies or jobs. Um, how do you, how do you, I mean, you're running one of the leading labs and how do you think about that public concern about the technology? [00:29:26] Demis Hassabis: I, I think, um, the public's right to be concerned. I think that there are things that, that, uh, and I'm concerned about several aspects of what the technology is, is a dual purpose technology. It's something this profound, you know, I sometimes describe it, quantify it as 10 times the, the, the, the impact the industrial revolution was, 10 times faster. And so, you know, you know, it's like taking a place over a decade instead of a century. So that's like a hundred X, uh, uh, of the industrial revolution. And it's probably an underestimate to be honest, but that's probably enough for us to try and comprehend, right? And deal with. And so, of course, uh, I think there are, uh, it's super exciting. There's going to be amazing things that are going to happen. Like we're trying to do that with solving all disease. I think, you know, a lot of the other challenges facing society today from climate to energy to, to, to disease will be enabled by AI. I'm sure of it. And in fact, I'd be much more worried about those challenges if I didn't think something like AI was coming down the line. Um, but it's going to cause a lot of change and, and disruptions and, um, and actually both on the technical side, economic and philosophical. And I think we've, we've got to, um, uh, think through very thoughtfully and bring together all parts of society to discuss this, not just the technologists. The technology and the safety of the technology is just one piece of this. Um, it needs economists, social scientists, human and, and humanity experts to kind of chart out what is going to happen next. And I think one of the reasons it's negative here is that, um, specifically because it's different in other countries. For example, when I go to, we came back from the summit in India, it's hugely popular with the youth of, uh, India because they see the opportunities that it's going to democratize for them. Having access to basically the same tools that you would have needed to go to Silicon Valley for, you know, the world. It's been an amazing moment in that everyone can access pretty much what's going on in the frontier labs, but only, you know, with a delay of just a few months, right? That's unheard of really, if you think about that. So these incredible things, but I think it's also partly the way some of my peers are articulating the, the, the, I think they're being very careful with their communication and understanding. They're being way too certain, I would say with some of their pronouncements where I think actually there's just huge uncertainty and that is, uh, worrying in itself. But it's also means that nothing is decided in my opinion. I, I think there's, it's, it's, it's kind of unknown. And I think anyone who says that I think directionally I could tell you some things, but I think a lot of it, it really depends on the actions taken, uh, in the next few years. And also what the youth of today, the many students in the room today, you're going to, uh, be forming. You're the first generation to kind of grow up in AI native. Should we say like I did computer native and just like every generation, you'll master these technologies become, um, super, uh, uh, uh, productive with them. And I actually think over the next 10 years, at least hard to predict beyond that, but you'll almost be super powered with those. And if you use them in the right way, the amount of creativity and projects you can do and the amount an individual will be able to do, but maybe that will change the nature of jobs. There'll be more entrepreneurial, small entrepreneurial things rather than big companies. I don't know. It's going to change a lot, but, um, and I think part of that is for society to come together and really take this exponential seriously. Um, not just the technologists, uh, economists and others that we were discussing this last night need to take this seriously right now and start charting out. What does that look like if we're in a post-scarcity world, for example, how does everyone benefit from that? You know, a lot of it's about how does, you know, it doesn't, it's obviously not correct for just a few people or a few companies or even a few companies or even a few nations to be benefiting from this technology. It needs to be broad. It's going to affect all of humanity has to broadly accrue the benefits to everyone, but how's that going to be done? We, we, and we really need, we've been, a lot of us have been talking about this for a while, but we really need answers now and concrete things and actions to be taken. And I plan to do my bit on that. I've been, um, thinking a lot about this over the years and planning and, and, and, and building sort of influence on this and I will do what I can. Um, obviously we're only, we're an important actor, but we're only one actor in this space. And I, but I hope, I know that the, the good news is I know that all of the sort of leading labs and the, and the leaders of those labs, although they disagree on a lot of things, they do worry about, uh, what, you know, these sorts of, um, issues coming down the line. But we need more forums to allow them to, you know, for us to come together to discuss these things, uh, more candidly. Uh, and I think that's what probably the public is detecting is just sort of slightly skewed discussions about what's going to happen. And, um, maybe for there's some ulterior motives behind some of that messaging, you know, raising money, other things. But I think we need to get to, you know, use the scientific method, be really, really rigorous and thoughtful about, uh, this, you know, critical moment in history. And then maybe the final thing I would say is I would love to see, I think it's incumbent on the industry and the field to show more unequivocally what the benefits are and, and not just talk about them, but demonstrate them. So in health, in medicine, in medicine, in science, um, these things are all, in my view, sort of unequivocal goods, right? Like alpha fold, but there aren't enough examples. There should be 20 alpha folds, right? And there should be, you know, we got to like stop talking sort of in the hypothetical about curing cancer and actually cure cancer. Right? And so this, these are the things that I think, um, are going to be needed to demonstrate to the public, you know, why, uh, are those of us who are excited about it? And many of us are in this room. Why are we excited about this? Why have we spent our whole life building towards this? And also how are we going to, uh, concretely mitigate the risks, um, while enabling all of the amazing things that we would like to see? And I think society needs. [00:35:16] John Levin: I think a lot, a lot of, a lot of great points there that if there were, if there were some tangible benefits that were realized because of AI breakthroughs, say to human health or drug discovery, that, that might change people's perception in some ways. And, and I love the, the suggestion of sort of trying to think farther out about a world that might look very different in terms of productivity and so forth. It's hard actually to do that. Yes. It's rare, you know, rarely in social science people can people get out of the current frame they're in and actually project way forward. I, I think of Keynes' great article during the depression when he looks out the economic lives of our grandchildren. It's a rare case that we, you were saying last night, you thought we needed another Keynes right now. Yeah. And that that's a, maybe, maybe, maybe there's someone in the audience who will do that. Let me ask, one of the things you've talked about for a lot of years is, is the need for the frontier labs to, to, in a sense, regulate themselves. That is to sometimes sort of not release certain kinds of technologies, you know, that might be threats to safety and so forth. Right now, it's pretty clear that the labs are just at a breakneck competition. They're investing everything. They're, they're going all out. There's, uh, do you still feel the labs ought to be self-regulating? Do you think the, you have to think the government ought to step in and regulate AI in some way? How do you see the current dynamic relative to the way you thought about, talked about it in the past? [00:36:54] Demis Hassabis: Yeah. Well, look, first of all, to, to, just to give some historical context to this, this is not what I, in terms of, we talked earlier about how the technology's gone. I think the technology has gone amazingly and maybe even on the, on the better side of what I imagined 20 years ago. But the, the, the environment it's kind of been birthing in is not the ideal. Far from it. Right? We were very worried about 15 years ago, 10 years ago about this race dynamic happening as more and more people, more and more companies, more and more ambitious, you know, uh, tech leaders, uh, realize what I had known for 20 plus years of how important this technology was going to be. Um, and we talked about some of this in the room with, um, about the dangers of this kind of, uh, race dynamic. And unfortunately we've ended up because of the way the technology's gone. So my, my, uh, what, if I could have waved a magic wand, what I would have done was, um, build AGI, the general technology more in a research, uh, uh, facility, perhaps like a CERN, maybe all the best minds helping critique each other's, uh, ideas and making sure we, we, we were rigorous with the scientific method and the testing of it and, uh, understanding each step that we took. Um, but then we wouldn't have to wait for that. Of course that means AGI would arrive, uh, uh, later, maybe 10 years later, but we wouldn't need to wait for that to get the benefits, societal benefits of it. Because at the same time we would break off bits of that and use it for specialized systems like more alpha folds curing diseases. That can be done, right? Because those are alpha fold is a specialized hybrid system, uses a lot of the ideas, the general, uh, uh, uh, uh, a purpose systems use, but it's specialized to protein folding. So that was actually my, and you can see that was my vision for it. Cause that's what we were doing. Um, but then chat bots changed that because, uh, effectively, and that was probably the only surprise to me of the last sort of 15 years on the science side is how effective transformers ended up being for language. And the fact that you could separate language and sort of learn it just from the internet without having to act in the world, either robotics or simulations, you know, it's kind of very interesting. And that'll be a whole another topic why that was, and I have some theories on that, you know, language is more grounded than we, than linguists probably thought. There's some, there's some grounding coming from the reinforcement learning feedback that the human testers are doing. Cause obviously we're, we're grounded in the real world. So when we say yes, no to certain things, our grounding is, is then ending up in a, in a, in a very low bandwidth way, but still ending up modifying what the, what the foundation model understands. So there was these sort of unexpected things I would say that happened. And then that made it a commercial, very important commercial technology that could be scaled with engineering and money and so on, which is what you see today. And that changed the dynamic and, and, and then has created what we see today, which is probably the most ferocious competitive environment. I would say there's ever been, I mean, certainly in the tech industry tech era, maybe ever, maybe other historians here from the business school will tell me otherwise, but it feels unbelievably intense being in the middle of it. And it feels like that for all of the participants. Um, and, um, and then on top of that, you layer the geopolitical complexities. So there's also, this is a double race going on. There's the, there's the, there's the, there's the race between the companies and it's pretty life or death for them. And then there's the, you know, us, China dynamic and, and others, right. The geopolitical down and there's a race there. So it's a double layered one, very tricky. Now, um, I, I still have hope that the, um, there can be some cooperation and coordination between. We certainly discuss this as lab leaders, uh, on the safety elements and the security elements. Everybody wants that. All nation, you know, nobody wants something catastrophic to go wrong. The problem is we're in a kind of prisoners dilemma where, where, you know, anyone who, by definition, if you take more time to release something or, or, or make it happen. Or, or, or make something safer. That's harder than just, uh, putting it out there and, and letting it see what happens. So you, so a defector has some advantage. Um, and that's the, this is the classic problem with the, the race to the bottom dynamic. And we've got to change that somehow. And I think urgently, and I think part of that is, um, some form of government, uh, involvement. Um, the hard part there, of course, is that anything to do with regulation, it's too slow. Like this is every week, there's something new. If we were to regulate something two years ago, it'd be just like, it's like ancient history now. So it would almost be, you know, almost certainly the wrong thing. So whatever is designed. And I have some ideas in this, and I'll probably be talking about this later this year is like, it needs to be dynamic, which is doesn't go. Usually that word doesn't go with regulation, right? So it's gotta be light, light, uh, uh, uh, fleet footed and able to informed by the, you know, latest developments, uh, so that it can adapt to where the actual risk is rather than some kind of perceived risk that turns out not to be the case or not the critical thing. Um, you know, many years before we just, it's just not going to work for AI. And, and, and even today that we, you know, the leading scientists wouldn't necessarily agree on a short list. In fact, I know they definitely wouldn't agree on a short list of what checks and balances are needed. So, and that's because the science isn't settled. It's, we're just, it's just, it's, and that's partly the speed. Um, but also the, the pace of the, the, the, um, progress, um, is, is running ahead of the understanding of it. That's just how it is. It's part of the, the race dynamic. Um, but, uh, uh, uh, we need to kind of somehow, uh, rebalance that. And I think, uh, some form of really almost, uh, smart regulation, uh, is required that is dynamic and can adapt with the times very quickly and probably informed by the, the lab, the leading labs. Uh, so, cause they're seeing what's actually at the, the cold face. [00:42:42] Speaker ?: All right. [00:42:44] John Levin: That, that, that open, I think that's a, there's so much more to discuss there in terms of the process. Uh, the, the, the, the aspects of how you set up a regulatory system for AI and, and do it in a way that didn't prevent some of the breakthroughs, positive breaks you're talking about. Exactly. [00:42:59] Demis Hassabis: To the geopolitics. We want all that innovation, right? We want to solve the disease. So exactly. How do you enable the good, the good use cases, um, you know, how, and without, and mitigate the bad. [00:43:08] John Levin: I'm, I'm looking forward to when you bring out your, your plan for that this year. I think that'll, that'll be fantastic. And that'll give us a lot to talk about here on campus. Yes. Next time. Uh, and everywhere. Just, uh, we've got a couple of student questions. I want to, I want to give a chance for some student questions. [00:43:26] Speaker ?: Cool. Uh. [00:43:28] Speaker 5: Hi, Demis. I'm Aaron's second year at the business school. My question is, how do you balance pushing the frontier of AI with ensuring that the health and scientific dividends, um, um, is like evenly distributed, um, in places like Africa and like the global south where the need, the need is like the greatest, but the infrastructure for like deployment and, um, research is most limited. [00:43:55] Demis Hassabis: Yeah. Yeah. We think about that a lot actually. And that, that was, that goes back to, um, one of some examples of that I can give is back to the alpha fold question. Where we, you know, folded all the, all the proteins. Um, we put that out on databases you could access from anywhere around the world. So these three million researchers come from 190 countries, just to be clear. It's pretty much every country, every researcher. Um, and it's, uh, it means that they, and what was great, what we did actually in the early days of seeding some collaborations, what you could do with alpha fold. We worked with, uh, we worked with, uh, the DNDI drugs for neglected disease, part of the WHO in, in, in Switzerland, which work on diseases in the poor places of the world that have, don't have good health. It has good health care systems. And some of those diseases are neglected, as you know, because, um, the, big pharma can't make money by, by, by, you know, uh, in those markets. So then the diseases that affect primarily those, those areas of the regions of the world, um, don't get as much research, uh, uh, resources behind them. So what we were able to do, uh, in collaboration with, with this institute and, and many actual universities on the ground, uh, is jump them straight to, not needing to trying to figure out like malaria virus or the structures of, you know, Zika virus or something like that. We should, they would have had to do all the painstaking structural biology. They can just start that as a given and work straight away on the drugs. So, um, you know, that allows to speed up massively the whole process. You know, they can sort of take the structure and, and, and of, of interest and, and move, move forwards from there. Same with crop resilience in, uh, affected by climate change. Um, we work with, uh, Jennifer Doudna's Institute and, and many others on, on these things. Because, um, lots of plant proteins, uh, you know, were not had, we didn't know what the structures were of them. Because obviously most of the structure work has gone into human proteins. So if it's animals or plants, there's a lot less, uh, data out there. So they were able to, so it's even more differentially impactful in those types of, uh, those types of areas. And then the final thing I would say is like, if we can, and, and I think this is where the capitalist engine can actually work for good here is. If we can make, um, the, uh, uh, the, the drug discovery platform that we're working on at isomorphic as efficient as I'm talking about. You know, down from years to months. So instead of it costing billions of dollars, it costs tens of millions of dollars. Maybe, maybe single millions of dollars. Then suddenly, um, what I'm hoping we'll be able to do with isomorphic is we, you know, we cure all these terrible diseases that maybe affect the, the, the, the richer parts of the world. That makes money and that fuels it, fuels the engine. But then we can do sort of philanthropically, the company could, uh, uh, uh, uh, find cures to diseases where we don't need to make any return. Because it's fast enough and it's cheap enough that it can just be done. Uh, and, and, and, and in a short amount of time. So I think that's sort of my dream for how I can make isomorphic help the whole world. [00:46:54] Speaker 6: Hi, Demise, thank you so much for taking the time to talk to us. My name is Miki, I'm a senior, um, in the Dorr School of Sustainability. And you've described extensively how AGI, um, could be humanity's most transformative technologies. And I'm just curious, um, the responsibility or how you think about the societal impacts alongside this intellectual pioneering, um, and productivity that AGI presents. Particularly when thinking about, you know, how is this going to redefine and reshape people, the challenges that we're trying to solve today, but kind of the downstream effects that that could bring forward. Thank you. [00:47:38] Demis Hassabis: Yeah, thanks for your question. I, I, I think about this all the time and, uh, have done from the beginning because, um, we were planning for success, right? So, uh, even though it seemed very improbable back, uh, 15, 20 years ago. And I think, um, this is why I, I, I like doing talks like this and, and meeting folks in, in these kinds of places is, I, I, it is a bit of a call to arms now. Like, it's very urgent that we, um, really think about the second order consequences. Um, and, um, I think many of you in the room and many of you in the humanity subjects, it's now's your time, in my opinion. Because, okay, we've got to get the technology right. But then, um, if we do that, then there's the economics question. And if we get in, and if we get that right, there's the philosophical questions about the human condition. And, um, and I'm very excited about, and I'm a big believer, I'm very optimistic. I should just very, obviously, I'm a cautious optimist is the way I would say it is. I, I, I'm, I'm very optimistic that we're going to get this right. And I'm a big believer in human ingenuity, uh, especially when the pressure's on. I think humanity's always figured it out when, when the, the chips are down and they are now. But we do really need to start taking that. I think the technologists are taking it seriously, but the other parts of society need to as well. Economists, I, you know, I'm, I'm always a little bit astounded when I talk to economists about what's happening. And it's sort of, they're pretty skeptical. Where's it, where's it coming in the GDP? And it's like, look, it's 10 times the industrial revolution. Can we start planning for that now? Like, you know, and, and we'll be in a world where, and we talked about this last night. I think we do need, uh, some giants of these fields like Keynes was, uh, for now. Like, why would that hold in a post-scarcity world? We're going to be in a world for the first time if we get the technology right, where we're a non-zero-sum world for the first time in humanity's existence. How can that not need a new type of economic system? Right? It has to. And I don't think it's any of the ones we've tried because they were all done under the guise of, um, zero-sum more, you know, and a limited, a scarce, uh, uh, uh, world. Um, and I'm talking about, you know, traveling to the stars and utilizing, um, all the, all the resources that are out in the solar system, not just the limited ones on Earth. And I think that really is going to happen if we get the technology right in the next 10, 20, 30 years. Um, and then after all of that, there's the even harder question of, of, of how do we want to evolve our society? And what is virtuous, what is meaning, what is purpose? And I think that's going to need lots of great philosophers. So, um, that would be my appeal to, to people in those fields is now, in my view, could not be more of an exciting time if you're working on those types of, of, of projects. As long, if, as long as you understand what's, and really viscerally sort of, uh, understand and lean into what's actually happening here. [00:50:29] John Levin: That's a good, good charge to university. Okay, one more student question. [00:50:32] Speaker 7: Hi, Dennis. I'm Janai. I'm a second year MBA student. My question to you is, what do you not want AI to touch in this lifetime? And what do you hold secret from your perspective? Thanks. [00:50:44] Demis Hassabis: Yeah, that's a great question. I, um, look, AI is going to be, in terms of the, the, the scientific world of things, um, it's a fully general technology, right? You can think of it as a, a Turing machine. That's the way I think about it. It was my favorite course at college. And, um, and I think our, our minds are actually fully general. So we're kind of approximate Turing machine. So as Turing showed, we, we, anything that's computable, a Turing machine can compute. And most things we know about in the universe, non-quantum things are computable. So that's a pretty large set of general things that we can turn our minds to. Hence, we built modern civilization, which is miraculous if we stop to think about it. And I don't think we wonder enough. We, we, we keep our sense, don't keep our sense of wonder for long enough about that. Um, but it also means these systems that we're building, they're also going to be sort of Turing powerful as well. Um, one thing I would say is, uh, that is, there are very big questions to come that I think it would be better if we took more time over. So one example that's pretty topical right now is consciousness. And we, you know, it's not a very well posed problem from philosophy and neuroscience still. Although I think we all have intuitions as to what are the important aspects of that. My feeling is the current systems are, are, uh, don't exhibit any, uh, are not, but others disagree. Um, we as, what I would recommend though, in terms of like what areas should AI not touch is that we build our first systems as tools, intelligent tools. That's enough of a challenge already in my opinion, cause that's already AGI. And then using those tools, I think we should study neuroscience and other things like that and, and philosophy and actually come up with a more rigorous definition of things like consciousness. I think that is possible. Um, and then test things against that. And then maybe society decide if we want to cross the second Rubicon of trying to make entities that at least seem like, uh, conscious to us. So we may not want to make that decision. I could write. I think that intelligence and consciousness are dissociable. So I don't think you have to do that to have an intelligent system. I think it's a choice. Some of our, uh, and so you can probably feel that in the, when you use some of the leading chat bots, um, there's differences in opinion that come through. Um, and my view is it'd be better to take that as two, as two steps. They're both enormous, um, for, for humanity, rather than conflate the two. [00:53:11] John Levin: Demis, we, you've, we've got an auditorium with many students in it. Uh, if you were back in school, um, how would you be thinking about, what would you be studying? How would you be thinking about what, you know, what, what should they, what would be your advice on how they should be thinking about to study in their careers? [00:53:29] Demis Hassabis: Uh, well, look, I would, I would be really excited, uh, if I was back at college now, um, my recommendation would be, um, those of you doing science and STEM subjects and mathematics and computer sciences still do those things. I think you'll be able to take better advantage of these tools if you understand how they are put together and what they're capable of. I think that's going to be true for the next, uh, the next period, the next 10 years at least. And, um, I would also lean in though to not wish it away. The genie's not going back in the bottle. Like, like lean into what these tools can do. I can tell you that the leading labs are so busy making the, the, the tools that we have not, we probably only scratched the surface of not even probably of what they can actually do. Even today's tools. I have this, you know, sometimes people call it capability overhang. There's so much potential these things can do if you figure out how to pair them with other things or compare it, compare it, pair it with another domain your expert in, um, build it into your workflow in an interesting way. You have those tools. They're the most powerful tools anyone's got. You have them in the palm of your hand. There's so much more you can do as an individual. I think it should unleash creativity. Like those of you studying humanities or, or product or business, maybe you didn't have coding, uh, uh, skills before, but you don't need, you don't, you can produce a lot of what's in your mind now, uh, using these tools. Right. Um, and, and I, but I think also the coders, the people who are expert that could do a hundred X more, like the terms of the size of project you can do if you're expert coding. So I think it enables both the democratization and the people that are specialized in those areas. So I think it's an amazing time, but it's also, I get, it's also worrying because everything's going to change. So that's the only thing I can tell you for sure. Everything is going to change in the next 10 years, probably more than people assume, but that also in any time there's enormous change like that. There's enormous opportunities that has to be right. And the world's sort of your oyster really. And I kind of envy some of you now cause you're the first generation that will be AI native, just like my generation was computer and internet native. And, uh, it's going to be in your hands in the end. Uh, the students in the room, like how that future world gets built. And, um, I think it's a very exciting time. If you think about it in, in the right way from the right angle and with a lot of imagination and creativity, but I think that's always been, uh, uh, true. Um, and maybe it's more so now in changes of, you know, periods of enormous change like this that accentuates it. [00:56:06] John Levin: We were saying last night that in a period of a lot of change where you don't quite know what the future holds, but you have to be able to be adaptable and have broad domain of knowledge. [00:56:17] Demis Hassabis: It's going to be a golden era for liberal, liberal education. So I think, I mean, the main thing is to just, um, make sure you double down your own agency. The future is still to be written. I would say that's so don't listen to anyone who says it's not. Yeah. [00:56:31] John Levin: Demis, thank you for joining us. It was amazing. [00:56:47] Speaker ?: Thank you. Thank you.