AI as the Ultimate Tool for Science: A Conversation with Demis Hassabis

Abstract and publication details

James Manyika. Demis Hassabis has focused on artificial intelligence from an early age. With a background in computer science and cognitive neuroscience, Demis first garnered wide attention for his work on AlphaGo, an AI system that in 2016 defeated a world champion in the complex strategy game of Go, long considered a grand challenge for machines playing games. But his ultimate goal was always much bigger: to use AI to solve problems in science and understand the deepest mysteries of the universe. His first proof point for this vision was AlphaFold, an AI system that cracked the fifty-year-old protein structure prediction problem. He was awarded the 2024 Nobel Prize in Chemistry and the 2023 Breakthrough Prize in Life Sciences for this work. Demis has also long been a pioneering entrepreneur in AI, most notably as cofounder and CEO of DeepMind, one of the AI labs leading the development of frontier foundation models. Fundamentally, Demis thinks of himself as a scientist, and so this dialogue focuses on AI and science: what motivates him as a scientist working with AI; what progress, opportunities, and challenges he sees; and more.Manyika. Why does science matter to you?Demis Hassabis. I’ve always been driven by a deep desire to better understand the world around us. The scientific method is arguably the most important idea humanity has ever had to make progress and advance knowledge. For me, there’s nothing more thrilling than being at the absolute frontier of what’s known, pushing the boundaries forward and discovering something new for the very first time-that’s science at its very best.Manyika. Why do you think of yourself as a scientist, and how did you get started as a scientist?Hassabis. Today, I think of myself first and foremost as a scientist. But when I started off my career programming AI for games and studying computer science as an undergraduate at Cambridge, I thought of myself as an engineer. Then as I followed up on my long-standing interest in the brain and started studying cognitive neuroscience, and specifically memory and imagination for my PhD, I realized I worked best in the mode of a scientist coming up with hypotheses about the world and experiments to prove or disprove them.But I’m also very practical and interested in real-world impact. That’s where the engineering half of my brain still comes in. Instead of getting lost in a theoretical argument, I lean into my pragmatic side and start thinking about how to build tools or experiments to prove hypotheses. AlphaGo and AlphaFold are good examples of that.1Manyika. What scientific accomplishments have inspired you the most?Hassabis. I’m a huge admirer and fan of the polymaths who lived during the Renaissance and Classical Greek eras, the greats like Aristotle and da Vinci, because they didn’t see any boundaries between subjects. I love that because it allows one to find connections between somewhat disparate domains. I believe those combinations are where the biggest advances will come from in the next ten to twenty years, and I think AI is going to help by finding commonalities or crossover between fields.I’m also inspired by enormous leaps of imagination and intuition that, looking back through the prism of time, come seemingly out of the blue. Einstein, Newton, Bohr, Ramanujan, and Feynman are some of my favorites-the scientists and mathematicians who have a mix of ideas, expertise, and interests going around in their minds and then make an intuitive leap in terms of how the world works.Manyika. What is on your list of the most important scientific problems that we haven’t solved yet?Hassabis. I got into science and AI because as a kid I was obsessed with, and I suppose in some way you could say haunted by, the biggest questions about the nature of reality and consciousness that underpin our understanding of the world. What is time? What is gravity? How does it all fit together with the quantum world? Amazingly, there’s not been a lot of progress on some of those fundamental questions in the last fifty-plus years, in part because they’re so complex.I’ve always seen AI as potentially the ultimate tool to advance the frontier of knowledge. Whether by processing data or assisting us in coming up with new conjectures, I think AI could help us tackle these fundamental questions.Manyika. How do you link these questions about the nature of reality with a computational approach to understanding them?Hassabis. I’ve always been fascinated by this question of what can and can’t be modeled by a classical computer or a Turing machine, and what the answer would tell us about the nature of reality. Most things in nature have a stable structure that evolved and survived over time-the behavior of a cell, the shape of a mountain, even the orbit of a planet. That suggests there should be some sort of pattern or process that is potentially learnable by a neural network. AI systems like AlphaFold are already showing us we can model many more complex systems with classical computers than we previously thought.Maybe in the limit most natural systems can be simulated or learned-kind of reverse engineered-by a classical system. That would support a conjecture I have that information is the most fundamental unit of physics, more so than matter and energy, and is the best way to understand the true nature of the universe.Manyika. In what way do you think information may be the most foundational aspect of understanding reality? If it is, how could a computational approach to understanding information help us answer the foundational questions?Hassabis. Traditionally in physics, energy and matter are regarded as primary. We can observe and measure them. Then people like Claude Shannon, Alan Turing, and John von Neumann-the founding fathers of computer science-introduced information entropy, measures of information, mutual information, and other important related concepts. But they came after the physicists, so they were trying to fit information theory within the existing framework of physics. Maybe it should have been the other way around, and physics should have been framed in terms of information.This notion also relates to Turing machines, of course, and what kinds of processes are computable on classical systems. Physicists assume that the universe is a quantum system, and that classical systems can’t model quantum systems due to their enormous complexity. But if it turns out a lot more physical processes in nature are actually able to be accurately modeled by classical systems, it may have some important implications for our understanding of physics.Ultimately, I’m most interested in the nature of reality, and I question whether our current understanding of quantum systems is the whole story. I suspect we might eventually find a simpler underlying description of what we are seeing that turns out to be more amenable to being modeled by a classical system, and I think we’re more likely to discover that if we think about the world as a computational universe.Manyika. Let’s take that a step further. If you take an informational or computability view of the world, don’t we run into difficulties, such as Kolmogorov complexity, the P equals NP question of whether problems are solvable in a tractable amount of time, or the Halting problem in Turing machines? Does that not potentially place a limit on how far we can go with the computational approach?Hassabis. It does, and I think those are the interesting limits to test and understand. P equals NP – which attempts to categorize the difficulty of a problem by how much computation it would take to find and check a solution, respectively – is one of the most important questions in science to resolve. I suspect P is not equal to NP, and there are some problems out there that are just not tractable to solve in a practical amount of time without invoking the help of, say, a quantum computer, but we need to understand this a lot better because there may be more nuance here than we previously realized. In our work with AlphaGo and AlphaFold, we’re showing that if you do a lot of precompute, which is not normally considered in these kinds of scenarios, you can seemingly answer some highly complex questions approximately optimally in P (polynomial) time. Neural networks are effectively using massive amounts of precompute to compress knowledge into some efficient artifact. That computed artifact is then available at test time and, for a lot of natural systems, you can use it to narrow down your search space so you don’t have to consider all the possible configurations they could potentially take, but only a much smaller subset that are actually plausible.Let’s take proteins. There are roughly 10300 possible conformations of an average protein. It would take longer than the age of the universe to enumerate that exhaustively to find the one specific shape it takes, so you have to do something much smarter. You have to learn what patterns there are for different amino acid sequences and then only search a tiny fraction of the possibilities to find the approximately correct solution. That seems to be what we managed to do with AlphaFold. Maybe not perfectly, but to an approximation that is at least good enough for practical purposes.Similarly, there are 10170 possible positions in the game of Go; there is no way to find the best move in a position through sheer brute force computation using simple heuristics. Instead, AlphaGo did this by learning a good model of plausible Go moves, so it only needed to consider the most fruitful paths of moves, and then from that much smaller number of options find the best move that would most likely get it to a winning position.There’s something very interesting going on here in successfully navigating these enormous search spaces, and my hunch is it’s pretty profound. We have a front row seat to these fundamental questions because of what we’re seeing in AI. Not just with AlphaGo, AlphaFold, and our science projects, but also with our video models like Veo 3. They are modeling some aspects of reality just by watching a few billion videos. It seems pretty amazing that they can reverse engineer some kind of intuitive physics: how liquids flow, reflections in glass, or the shadows cast by objects.All of this is telling us something about the underlying nature of physics, including what the limit of a Turing machine is versus what you need a quantum system for, and, therefore, something about the nature of reality itself. AI has implications for all those questions.Manyika. AlphaFold seems to be a great example of what you’ve said about the natural world having structure that can be modeled or learned by neural networks. Talk about how you did that in the case of AlphaFold.Hassabis. AlphaFold was our solution to the protein folding or protein structure prediction problem. You start with an amino acid sequence-you can think of it very roughly as the genetic sequence for the protein, a one-dimensional string of letters. In the body or in nature, that string folds up into a 3D structure, and that shape goes a long way toward defining the function of that protein, which is really important for drug discovery and disease understanding.Predicting those structures is a fifty-year-old grand challenge in biology. We managed to solve it with AlphaFold, at least to an atomic accuracy, which is the level of precision needed for experimentalists to find it useful.The way we did it is that there were about 150,000 known structures that had been painstakingly put together by structural biologists over the past thirty to forty years with very expensive equipment like electron microscopes. That was just about enough data to give our AI system clues as to the topology of proteins. Of course they don’t just fold up randomly; there are some constraints, and the AI system learned them. Eventually it was able, within a few seconds, to come up with a plausible structure for an unseen protein. Then over the course of a year we folded all two hundred million proteins known to science and, with our colleagues at the European Bioinformatics Institute, made them freely accessible in a vast database for researchers all around the world to use.Manyika. Of course, you’ve extended this to AlphaFold 3 and actual protein interactions. Are cells next?Hassabis. AlphaFold 2 essentially predicted the static snapshot structure of a protein. But we know biology is a dynamic system and we need to understand protein interactions and how they change their shape and behavior.AlphaFold 3 is our next step on the ladder.2 We’re now predicting all pairwise interactions between proteins and proteins, proteins and ligands, proteins and RNA, proteins and DNA. Basically, all of life’s molecules and their interactions in pairs can be modeled by AlphaFold 3.But to scale up to an entire cell, we’re going to need to do much better than that. There are cascades of interactions and more than two biomolecules are involved. It is obviously hugely complex. The next stage might be modeling a pathway and then maybe a simple cell, like a yeast cell, which is an entire organism. Then, if that works, maybe we simulate an organelle from a more complex organism, and then eventually an entire organism.I’ve had this dream for over twenty years. The idea would be to have a realistic enough simulation of a cell that you could do in silico experiments maybe orders of magnitude faster than you could do in the wet lab. Then the lab is saved for the validation step and the grounding of the results rather than for the exploration phase. A virtual cell would be an incredible boon not only for fundamental biology but also for drug You have also on other in science – physics, systems, and so the there that these are all natural systems, and they to being We for aspects of a problem in whether it is to tackle with the AI we have can the problem be as or into a description of a massive it’s and brute force work. if do you have enough data to learn some sort of model of the topology of that maybe a is available or learnable that can some you have you need a that trying to or In that is winning or the In a natural system, that might be the energy in that system. If you can that, you can then use a model to search with the of the function toward the solution. It turns out a lot of problems fit that that, we for problems where the of it would be If you think about the of all there are some problems that, if you solve you entire new of folding was one and it up in drug discovery and disease Whether for energy, systems, or fundamental of the problems we to work on have those more about It is not that all of has structure, at least not structure we understand. of is a good obviously still an question about how much structure there is in I suspect it is a lot because was to the world, which does have structure, and it seems to be able to do that very are mathematicians like Ramanujan, who was intuitive and came up with amazing without going through all the so much so that he these leaps of imagination were We it knowledge or intuition because we can’t it or it in a intuition is not it is just the we use for the knowledge the brain through who are very good at say, like have efficient and highly of that knowledge. still a process there and I think in the limit it could be modeled by an AI AI to and already seeing a of what can be in the which is a challenge for the in the The model is a problem that it into a then it in an way by some on the current working of a proof it a correct different from understanding if there are patterns in the or structure in That would kind of system. I think are interesting – AI and finding patterns in I think may be What and do you see in current AI systems when it comes to being able to do What new do you think are and one of the I don’t think systems are to is that they don’t have true What I by that is, can you have an AI system that just solve a which is obviously and but can actually come up with a or conjecture like the Prize problems or the I think much more the question is the part of the and science We don’t understand how that level of in and we don’t have it in our AI other in current AI systems is the one these systems can solve very on the other if you the question it can still up on That be the case for a They would have of the What does the idea of to I’ve always as a system that all the cognitive the That’s a point because the brain is the only proof that we know of so in the that intelligence is even The brain is and incredible that with minds evolved for we were able to the scientific method and build the of around from to to The of great and all from this brain In my we really know for that a system we’re able to the of what we know that brain can a theoretical point of this is related to the of a Turing machine, which Alan Turing could any time and The brain can be as a approximation to a Turing machine, that, in it could learn that is as with any system, there always has to be some of because there is only so much time, and practical system can be for possible we can’t the The is the to learn as we are able to and build tools and machines to help us system should have the the to the incredible of including being of true intuitive leaps of discovery and coming up with a or could be by, for an AI system with a knowledge of and seeing if it is able to come up with like did in For the answer is foundation models can already do is and in my there’s no that they will the of any system. Whether or not up these models is enough on its to is an and I would not be if a number of on the level of or deep learning were still at we to forward the of and in In whether on the one one could use a of and, on the other there is some intuitive leap or that will like move AlphaGo strategy and intuition to an move that to over world champion Go say, I think got it or is it one or the I think you need is such an important in and I think we should be really scientific and about There should be a lot of examples of or where some new physics is or some new conjecture is that mathematicians is Instead of just coming up with a great new strategy in Go like AlphaGo could it a game as as I would to see examples of this for it to at least start even being a I think we should have this scientific approach to make there any or in the system. example of that is very for the average to come up with things they can’t For they can’t even to an level or learn a new and they don’t There are problems with and of those would need to be solved in to a test like the one I was How should scientists think about AI when it comes to AI a a something how do you the of scientists in the when we have these more For the next or we should think about AI as this amazing tool to help whether number pattern or some sort of or in these I think it is going to be the ultimate tool to help advance science and and it will in a new of and hypotheses are going to come from the and then have these tools that make them much more ten or one hundred what they that it is to say with any but these systems will more like What does it to you to use AI in science I know you about but what does that actually It to the scientific method and being and with your and the made about the systems. It also being very about what you and use the systems for and trying to think through their going to get that because it’s such a new and the frontier is so but one should make possible to that, it using the scientific method to help with understanding AI its its and its That understanding should then be to put the on AI, which are then and as think the scientific method is the best approach to the challenges with AI as it more and more rather than a more of first and questions That has its of course, and it’s amazing for progress, but with something as and as as AI, we need to use the best tools we which in my is the scientific to have as much and understanding of these as we What are the biggest to humanity that like to see from your one is and those that the of the world. I think we can do so much better than we’re there is with energy, and the by new and existing and and also modeling what’s to the world and how we’re In my those are the most important things we can use AI for in the If we’re having this in what are we going to be the of how things are by the only we know for is that by things will likely very the will have into the world for the of all and we will be how to a world of and in a new of scientific discovery and

Linked researcher profiles