The Quest to ‘Solve All Diseases’ with AI: Isomorphic Labs’ Max Jaderberg

After pioneering reinforcement learning breakthroughs at DeepMind with Capture the Flag and AlphaStar, Max Jaderberg aims to revolutionize drug discovery with AI as Chief AI Officer of Isomorphic Labs, which was spun out of DeepMind. He discusses how AlphaFold 3's diffusion-based architecture enables unprecedented understanding of molecular interactions, and why we're approaching a "Move 37 moment" in AI-powered drug design where models will surpass human intuition. Max shares his vision for general AI models that can solve all diseases, and the importance of developing agents that can learn to search through the whole potential design space. Hosted by Stephanie Zhan, Sequoia capital Mentioned in this episode: Playing Atari with Deep Reinforcement Learning: Seminal 2013 paper on Reinforcement Learning Capture the Flag: 2019 DeepMind paper on the emergence of cooperative agents AlphaStar: 2019 DeepMind paper on attaining grandmaster level in StarCraft II using multi-agent RL AlphaFold Server: Web interface for AlphaFold 3 model for non-commercial academic use

• 0:00Speaker 00

  We have set up the company from day one to really go after this big ambition. This isn't about developing therapeutics for a particular indication or a particular target. It's really thinking about how do we create a very general drug design engine with AI, something that we can apply to not just a single target or even a single modality, but we can apply this again and again across any different disease area. And that's what we're stepping towards at the moment.

• 0:48Speaker 01

  Today, we're excited to welcome Max Yatterberg to the show, Chief AI Officer of Isomorphic Labs, which launched out of DeepMind with the goal of revolutionizing drug discovery using AI. Last summer, they released AlphaFold 3, a stunning breakthrough that allows us to model not just the structure of proteins, but of all molecules and their interactions with each other. That led to Demis Hassabis winning the Nobel Prize in Chemistry last year. Max describes their vision for what a holy grail model for drug design and what agents for science look like. He draws parallels to his experiences building AlphaStar and Capture the Flag and the research directions of building agents and games more broadly.

• 1:30Speaker 01

  Specifically, with 10 to the power of 60 possible drug molecule structures, we need to build both generative models and agents that can learn how to explore and search through the whole potential design space. Max also describes his vision for what a GPT-3 moment for the field might look like, describing it more akin to AlphaGo's famous move 37 when we start to see things that exhibit superhuman levels of creativity in AI drug design, and that stun even humans ourselves. This is one of my favorite episodes yet. Enjoy the show.

• 2:05Speaker 02

  Max, thank you so much for joining us today here in London.

• 2:08Speaker 00

  Nice, a pleasure to be with you here. Yeah, it's fantastic.

• 2:12Speaker 02

  Awesome timing too with the launch of AlphaFold 3 and with Demis winning the Nobel Prize in Chemistry, which is a true testament to everything that you and your team have done over the last couple of years.

• 2:21Speaker 00

  Yeah, 2024 was definitely a busy year for us. Lots of big breakthroughs. Nobel Prize was just incredible to see. You know, I think amazing recognition for this seminal piece of work.

• 2:32Speaker 02

  Yes. Well, I'd love to start with talking a little bit about your own personal story. You've had an incredible career in the world of deep learning from the very start,

• 2:42Speaker 02

  authoring many seminal papers while at DeepMind, including for Capture the Flag and AlphaStar, breakthroughs in the world of deep learning. Can you walk us through some of the key questions that you had in your field of research around reinforcement learning at the time?

• 2:59Speaker 00

  Yes, so at DeepMind, I worked on a whole host of stuff, you know, early days of computer vision and deep generative models, but it was really reinforcement learning that ended up hooking me there. DeepMind was the place in the world to be working on reinforcement learning at that time. And, you know, really the question in our minds was, how can we actually get to a point where we could get an AI that could go off and do any task you wanted it to do? And, you know, the dominant paradigm at that point in time was supervised learning.

• 3:37Speaker 02

  Yeah.

• 3:38Speaker 00

  And supervised learning is very different from reinforcement learning. They're both learning techniques, but supervised learning, you need to know what the answer to your question is, and that's how you train the model. So in supervised learning, you give an example and then you supply the model with the answer to that question.

• 3:57Speaker 00

  Now, that can be great if you already know everything about the problem that you're training this AI to do, this neural network to do.

• 4:05Speaker 02

  But most times you don't.

• 4:06Speaker 00

  Yeah, I mean, there's, there's just so many problems in the world where we don't know what the answer is. Uh, we don't know what the solution is. And if you think about, you know, I think about how I want AI to be applied to the world. Yes, it's going to be great to be able to apply things where we're already good as humans here. But really, you know, the, the, the big frontier is can we start applying AI to places where, you know, humans don't know how to do this stuff, or, you know, there's a limit to human performance there. And, um, you know, that's where reinforcement learning is one of the key tools and yeah, has real promise here. Because in reinforcement learning, you don't need to know what the answer to the question is.

• 4:47Speaker 00

  You just need to be able to say whether the answer that the model gave you was good or not good.

• 4:54Speaker 02

  Yeah.

• 4:54Speaker 00

  Maybe even how good or not good. And, um, so this opens up a completely new, you know,

• 5:02Speaker 00

  field of problems to train these models against. And so reinforcement learning and, and really, you know, starting from what was one of the big breakthroughs of DeepMind in, in the early days was, was, you know, working on games like Atari.

• 5:16Speaker 02

  Yes.

• 5:16Speaker 00

  The question was, okay, so how can we scale this up from, you know, the world of Pong and Space Invaders to things that really start to look like real problems in the world?

• 5:27Speaker 02

  Yeah.

• 5:27Speaker 00

  Um, and so there was amazing track of research, um, as we scaled up these methods. Yeah.

• 5:34Speaker 02

  Did you know that Sequoia was, uh, the first investor in Atari back in the day?

• 5:38Speaker 00

  Oh, really? I didn't, I didn't know that. That, that, that's incredible. Yeah, no, those, those Atari games were, you know, great fun actually to, um, to sort of go back and play in the context of, hey, we've got an agent and, you know, I'm just going to have a game of Pong on the side as well.

• 5:53Speaker 02

  There's a wonderful wall at Sequoia in our office where we have all these names of legendary IPOs and M&As that have happened. And there's one, I think it's called The Pizza Company. And I love asking folks if they know what that is, and it's actually from Chuck E. Cheese's, which was an original Sequoia investment at the time.

• 6:13Speaker 00

  Amazing, amazing.

• 6:15Speaker 02

  So Capture the Flag and AlphaStar were incredible breakthroughs at the time. Can you share a little bit about what exactly those breakthroughs were? And maybe why you chose those specific games?

• 6:28Speaker 00

  Yeah, so, you know, if you think about the history of AI using video games, why do we use video games at all? Video games are these sort of malleable, perfectly encapsulated worlds that as researchers and scientists, you know, we can manipulate them. We can test out different algorithms in them. We can set up different situations. So the perfect test ground for us to develop new algorithms. And then you can imagine as a RL researcher, as someone who's like thinking about how can we get AI to be as general as possible? You're always thinking, okay, we've cracked Atari. How do we get a more complex game?

• 7:08Speaker 02

  Yeah.

• 7:10Speaker 00

  And the thing that I was personally obsessed with is I want these agents to be able to zero-shot be able to do any task. And this is a slightly different paradigm from the, from what people were doing at the time with training on Atari where, you know, normally in reinforcement learning, you think about, here's a game. Now you get to train on it and get good at it. And then you apply that same algorithm from scratch, training on different games.

• 7:37Speaker 02

  Yes.

• 7:39Speaker 00

  I'd love a different scenario where instead we train an agent and then we can lift it and put it on any new task.

• 7:46Speaker 02

  Yeah.

• 7:46Speaker 00

  And that agent will be able to perform well in that task without any more training.

• 7:50Speaker 02

  Yeah.

• 7:52Speaker 00

  And so to do that, what you're really asking for is generalisation over task space.

• 7:56Speaker 02

  Yes.

• 7:58Speaker 00

  And that means you need lots and lots of training tasks. So the training data in this RL for agents becomes tasks.

• 8:07Speaker 02

  Yeah.

• 8:08Speaker 00

  Not images, not pieces of text, but tasks. And so you can imagine you could go and sit and, um, you know, take a whole game studio and try and hand author, you know, hundreds of different tasks, lots of little mini games in these virtual worlds. Um, and we did that, you know, we, we, we were doing lots of that. And then you can think, yeah, we can actually go further than hand authoring. We can procedurally generate.

• 8:31Speaker 02

  Yeah.

• 8:32Speaker 00

  Um, these tasks and games, uh, generating worlds and, and, and, um, maps and, and, and different objectives. And we did that. But you keep running into this complexity ceiling that there's only so much complexity that you can hand author, or you can design.

• 8:50Speaker 02

  Yeah.

• 8:50Speaker 00

  Humanly.

• 8:52Speaker 00

  But that's where multiplayer games come in.

• 8:55Speaker 02

  Yeah.

• 8:55Speaker 00

  Because as soon as you go from single player to multiplayer, it's not just the agent playing. You've got another player in this game. Um, and that other player or other players can take on many different characteristics and many different behaviours. So every different player, every different strategy that you're up against changes fundamentally the game and what the agent is trying to do. You know, I, I, I go back and think, what, you know, why are people still obsessed with playing chess?

• 9:23Speaker 02

  Yeah.

• 9:23Speaker 00

  Why does a professional chess player still keep playing chess? It's the same game.

• 9:27Speaker 02

  Yeah.

• 9:28Speaker 00

  But it's actually not because you're playing completely different opponents day after day and new people into the world. Um, so the game is continually changing. So multiplayer games and multi-agent games really encapsulates that huge diversity of tasks that you might encounter just from other players, um, being there. And so Capture the Flag was actually one of our first forays into how can we use multiplayer games to really stretch what our reinforcement learning algorithms can do to really force us to think strongly about how we can generalize to new tasks, how we deal with these multi-agent dynamics. So Capture the Flag was a fantastic breakthrough, really showed that we could get to human level performance

• 10:14Speaker 00

  for these multiplayer first-person games.

• 10:16Speaker 02

  Yeah.

• 10:17Speaker 00

  And then, of course, StarCraft added on a huge amount of complexity and was sort of the next frontier that we had to go after for this.

• 10:25Speaker 02

  You were so early in this, so many of these concepts are very, very relevant today in the world of language. How does it feel to to see some of this work continue to be played out?

• 10:35Speaker 00

  Yes, brilliant. It's it's it's just fantastic, actually. You know, there were so many things that we were talking about.

• 10:41Speaker 02

  Seven years ago, yeah.

• 10:42Speaker 00

  Yeah, yeah, you know, twenty fifteen, sixteen, seventeen, eighteen. And to see all of these core fundamental concepts be really useful and really applicable today in the world of large language models. You know, and and resulting in performance that we could only really dream about at the time, that's that's incredibly satisfying.

• 11:02Speaker 02

  Yeah. So then, you know, in your own words, you said that you moved from building toys to then finding real applications. When did you know that you found the right recipe?

• 11:15Speaker 00

  So, you know, I I I just love deep learning. I've been obsessed with deep learning for, you know, ten, fifteen years now. And the thing that I love about it is that you have these underlying core concepts, these fundamental building blocks that are somehow incredibly transferable between different application spaces.

• 11:39Speaker 02

  Yes.

• 11:41Speaker 00

  So, you know, it's the same building blocks that we were using in computer vision in twenty twelve as we were using in, you know, generative, early generative models in language, you know, then reinforcement learning, et cetera, et cetera.

• 11:54Speaker 00

  So what I was seeing just again and again was this ability to take these core concepts, these same core concepts.

• 12:04Speaker 00

  Take incredible people um who understand how do you, you know, they're almost like master chefs of putting these these concepts together and these different building blocks together. Take a team of incredible people um and go after, you know, really, really challenging problems. You know, problems that you go to conferences at the time and you and you talk to leading researchers in the field, they say, no, no, no, this is ten years away. And in the back of your mind, you know, okay, we actually, we basically cracked it.

• 12:33Speaker 02

  Wow.

• 12:34Speaker 00

  And I saw I saw that happen again and again and again. You know, you take amazing people, amazing algorithms, amazing computes on really challenging problems and we can find recipes now to crack so many problems. And so it it just got to the point where, um and I've always been quite obsessed with the application of these methods. I I wanna see this technology have, you know, real transformative, positive impacts in the world.

• 13:02Speaker 02

  Yeah.

• 13:03Speaker 00

  Uh, and so, you know, we need to start actually going after that.

• 13:07Speaker 02

  Yeah.

• 13:07Speaker 00

  And and the, you know, the the time has been right for, I think, a few years now.

• 13:11Speaker 02

  Yeah. Well, so you've now had a decade-long relationship working together with one of the greatest scientists, technologists and founders of our lifetime, Demis. He called you while you were still at Oxford. And then your company, Vision Factory and DeepMind, were both acquired by Google back in twenty fourteen, around the same time. And that's when the two of you started to work together um now for over ten years. What was it like or what has it been like to work with Demis?

• 13:42Speaker 00

  Yeah, I mean, Demis is a incredible person. You know, a a real character and a real visionary.

• 13:50Speaker 02

  Yeah.

• 13:52Speaker 00

  And, you know, also amazingly human and relatable. And I think that that really inspires people. So, you know, it it only takes fi- a five-minute conversation to, for him to sort of really bleed out the the depth of ambition that he thinks about. And just the the immediacy of of of of the potential to get, you know, to step towards these ambitions. So I think um.

• 14:22Speaker 00

  He he has this great ability to inject a lot of energy into, you know, a group of very smart people.

• 14:29Speaker 02

  Hmm.

• 14:29Speaker 00

  Um, get people to see beyond what's right in front of them.

• 14:33Speaker 02

  Hmm.

• 14:34Speaker 00

  Uh, you know, I remember moments sitting, uh, well, standing in the lobby of one of the early DeepMind offices. I think this was the, it was a toast. We were a celebration we were having for the first nature paper from DeepMind.

• 14:47Speaker 02

  Wow.

• 14:48Speaker 00

  And, um, Demis was saying, you know, this is actually just gonna be the first of dozens of nature papers.

• 14:56Speaker 02

  (laughing).

• 14:57Speaker 00

  And at the time, this was the first, basically the first machine learning paper in nature. This was the Atari DQN paper. And the prospect of dozens of nature papers, you know, it seems a bit farfetched. And actually, he went further and said, and and and we're gonna be winning no more prizes as a result of this. And that was ten years ago.

• 15:17Speaker 02

  Yeah, that's incredible.

• 15:18Speaker 00

  But the forethought that he has, um, he's got what I call like one of these rollout minds. Maybe it comes from all of his experience playing chess, but it's he's always, you know, rolling out into the future.

• 15:29Speaker 02

  Ah, interesting.

• 15:29Speaker 00

  What what are the steps now that are gonna lead, you know, to this big ambition? Um, so yeah, it's been it's been fantastic. I've been working with him for about ten years now. Um, you know, still work really closely together on Isomorphic Labs. Um, and the ambition is as big as ever.

• 15:46Speaker 02

  It's so interesting to hear that, um, you had this ambition and that he had this ambition from the very start. Um, and it's incredible that it's played out that way.

• 15:54Speaker 00

  Yeah.

• 15:55Speaker 02

  Um, well, I'd love to talk a little bit about Isomorphic. You're now embarking on one of the most ambitious missions of our generation to reimagine drug discovery and drug development with AI. If everything goes right and you realize your vision for Isomorphic, what does the world look like?

• 16:13Speaker 00

  Yeah, you know, we we think really big isomorphic.

• 16:17Speaker 00

  We want to be solving all diseases here and and and genuinely that that that scale. And the point is that this technology that we're building, you know, and AI as a whole field is going to be completely transformative in how we understand biology. In our ability to manipulate and craft chemistry to modulate that biology. So we really think about a future where we are solving all diseases, where, you know, AI is not just helping us, you know, discover and create and design new therapeutics. But also just understand so much more about our biological world, about how our, you know, cells are working.

• 17:01Speaker 00

  And what are the root causes of of disease? And therefore, opening up new pathways that we can think about modulating.

• 17:08Speaker 02

  Yeah.

• 17:09Speaker 00

  So we we have set up the company from day one to really go after this big ambition. This isn't about developing therapeutics for a particular indication or particular targets. It's really thinking about how do we create a very general drug design engine with AI? Something that we can apply to not just a single target or even a single modality. But we can apply this again and again across any different disease area. And that's what we're stepping towards in the moment.

• 17:44Speaker 02

  How does setting out with this ambition of being general change how you built in practice from day one?

• 17:54Speaker 00

  Yes, a good question.

• 17:57Speaker 00

  When I think about some of the status quo of AI in drug design, there's a lot of. There's been a lot of use of machine learning models in chemistry and biology, but I would call them, you know, a lot of the first generation of this sort of application to be more local models.

• 18:15Speaker 02

  Yes.

• 18:15Speaker 00

  Where you might have some data about a particular target or about how a particular class of molecules is behaving and you'll fit a small, you know, multi-layer MLP against this data.

• 18:28Speaker 02

  Yeah.

• 18:30Speaker 00

  To help you generate some predictions that lead to your next round of design.

• 18:34Speaker 02

  Yeah.

• 18:36Speaker 00

  This is the complete opposite approach of what we were trying to do. So from day one, we were setting out to create models that generalize across chemistry and across target space. So, you know, and a key example of this is something like AlphaFold and AlphaFold 3.

• 18:53Speaker 02

  Yeah.

• 18:54Speaker 00

  Where this is a model that you can apply to a whole different host of targets. You can apply it to any protein in the proteome, in the universe of proteins. You can apply it to any small molecule that you can think of designing without needing to fine tune it, without needing to fit any local data.

• 19:14Speaker 02

  Yeah.

• 19:14Speaker 00

  And so you can, you can imagine that completely changes the way that chemists can use these models if you don't need to be adapting this model to every single application.

• 19:22Speaker 02

  Yeah.

• 19:23Speaker 00

  So every single one of our internal research projects, and by the way, when I think about, you know, what we're going to need to get this breakthrough drug design engine that we've been building, we need like half a dozen AlphaFolds.

• 19:37Speaker 02

  Wow.

• 19:38Speaker 00

  AlphaFold is just part of the story.

• 19:40Speaker 02

  Wow.

• 19:41Speaker 00

  So from day one, we've been setting up these internal research programs, going after these half dozen problems. We've had significant breakthroughs, obviously in AlphaFold and structure prediction, but also in other key areas.

• 19:56Speaker 00

  And in all of these, these models are general. They can be applied to any target. They can be applied. And then, and then what we're finding actually, they can be applied to any modality or lots of different modalities.

• 20:07Speaker 02

  Yeah. So that's the first time I've, I've heard you say half a dozen alpha folds. Can you share a little bit more about what that means?

• 20:15Speaker 00

  Yeah, so, yeah, alpha fold was obviously a massive breakthrough in, in understanding

• 20:21Speaker 00

  biomolecular structure. So how, what, what is the structure of proteins? And now with alpha three's structure of proteins with small molecules and things like DNA and RNA. That's a fundamental step change. It allows us to get experimental level accuracy of a really core concept of biochemistry.

• 20:38Speaker 02

  Yeah.

• 20:38Speaker 00

  That unlocks a whole bunch of thinking and design work for chemists.

• 20:41Speaker 02

  Yeah.

• 20:43Speaker 00

  But, you know, my, my comment here is actually, we're probably gonna need something like half a dozen more of these sort of breakthroughs. This sort of getting to experimental level accuracy

• 20:52Speaker 02

  Yeah.

• 20:52Speaker 00

  of different core concepts of biology and chemistry.

• 20:55Speaker 02

  Yeah.

• 20:56Speaker 00

  To be able to put this together into something that's really transformative for drug design. Drug design is really, really hard.

• 21:04Speaker 02

  Yeah.

• 21:05Speaker 00

  It's not just a single problem.

• 21:06Speaker 02

  Yeah.

• 21:06Speaker 00

  It's not just about understanding the structure of a protein.

• 21:09Speaker 02

  Yes.

• 21:10Speaker 00

  Um, it's not even just about designing a molecule that will modulate that protein.

• 21:15Speaker 02

  Yeah.

• 21:15Speaker 00

  In the way that you want.

• 21:16Speaker 02

  Yeah.

• 21:17Speaker 00

  You want this molecule to be able to, you know, ideally be taken as a pill and go through the body and be absorbed in the right way and reach the right, you know, cell type. And actually, you know, go into the cell and, and, and, and, and not be broken down by the liver in a certain way.

• 21:30Speaker 02

  Yeah.

• 21:31Speaker 00

  There's just so much complexity.

• 21:32Speaker 02

  Yeah.

• 21:33Speaker 00

  To hold on to as a drug designer.

• 21:35Speaker 02

  Yeah.

• 21:35Speaker 00

  Um, and each one of those is like, you know, an alpha fold level

• 21:38Speaker 02

  Yeah.

• 21:39Speaker 00

  style breakthrough that, that we've been creating.

• 21:41Speaker 02

  So interesting. Um, well, I've also heard you use the words a holy grail model for drug design and agents for science. Can you explain a little bit more about what you mean?

• 21:51Speaker 00

  Yeah, so, um, some of these research areas that we've been going after,

• 21:57Speaker 00

  um, you know, predicting structure and properties of these molecules and how all of these like biomolecules interact and play out over time. Um, these really are sort of holy grail predictive problems for drug design. Um, and we've made some incredible breakthroughs there, which have, you know, really stunned our chemists and, and, and step changed how we do drug design internally, ISO. Um, but what's, I think, a really interesting thing to think about is that.

• 22:26Speaker 00

  You could be, you could create the best possible predictive model of the world, like an experimental level, even better than experimental level model, to predict a particular property about a molecule, for example, to be able to predict the outcome of a real experiment. See, we could have a whole suite of those, but that still wouldn't solve drug design. And, and the way to think about this is, is, you know, there's this number 10 to the power of 60, which is perhaps all of the possible drug-like molecules that you could, that could exist. Wow. You know, that's maybe, that's maybe, that's maybe, you know, a bit, you know, takes into account a lot of things. So we could even reduce that by 20 orders of magnitude.

• 23:11Speaker 00

  Yeah. Get to 10 to the 40. Yeah. That's still a lot of things. Yeah. And even if you had the best predictive models in the world, so let's say you could screen a billion different molecules, you could go and test a billion different molecules, that's 10 to the 9. So, you know, now we're still like 10 to the 31 molecules left on the table. Yeah. So even with the best predictive models, you're still not even scratching the surface of molecular space that you should be exploring. And this is why we need to go beyond just predictive models of experiment, but also models, like generative models, like agents that can actually navigate that whole 10 to the 40, 10 to the 60 space. That's so interesting.

• 23:53Speaker 00

  Using our predictive models, obviously, to understand how to navigate that, but so we don't have to exhaustively search, because we can never exhaustively search the whole universe of molecules, if that makes sense. Just in the same way that AlphaGo couldn't exhaustively search. Right. All of the possible Go moves. Right. Unlike chess, where you could exhaustively search all possible chess moves. Yeah, yeah. But, yeah, molecule design is much more like Go than it is like chess. So that's where generative models come into play. Agents that utilize generative models, utilize search techniques as well as these amazing predictive capabilities to really open up the entirety of molecular space.

• 24:34Speaker 00

  Now, to me, it's actually still amazing that even without AI, we managed to find drugs in this ten to sixty space.

• 24:41Speaker 02

  Yeah.

• 24:42Speaker 00

  Ten to the forty space.

• 24:43Speaker 02

  Yeah.

• 24:44Speaker 00

  It just says, actually, there's probably a lot of redundancy. There's a lot of potential designs.

• 24:49Speaker 02

  Yeah.

• 24:50Speaker 00

  You know, if you think about a particular disease indication, a particular target, then there should be many designs that exist that would be good for that and would be the right sort of product profile for this therapeutic.

• 25:03Speaker 00

  And I think the real potential here is for these generative models, these agents as well, to be able to search through this space and really uncover that whole potential design space.

• 25:14Speaker 02

  That's so interesting. I think in very simplistic Lehman terms, you're both modeling, learning and modeling the game and trying to build the best player to solve different types of games.

• 25:28Speaker 00

  Yeah. So, I mean, you know, I'm incredibly biased by games. I've been playing video games since I was a kid, grew up in that world. But, you know, that's exactly how I think about it. We've got to be creating our world models, our models of the biochemical world, our biological world.

• 25:49Speaker 00

  And then we don't stop there. We actually then need to be creating agents and generative models that can work out how to explore,

• 25:55Speaker 02

  Yeah.

• 25:55Speaker 00

  how to traverse that and to basically uncover these, you know, amazing needles in the haystack in chemical space, which could be, you know, life-changing therapeutics for so many millions of people.

• 26:06Speaker 02

  I love that. That is our punchline today. So Allopaul 3 is truly groundbreaking. You've taken us from being able to model just the structure of a protein to now being able to model the structure of all molecules and their interactions with each other. Can you share a little bit about how we should think about that in terms of the impact in accuracy, in speed and efficiency, and also potentially in being able to explore problem spaces that we couldn't solve before this?

• 26:37Speaker 00

  Yeah, so.

• 26:39Speaker 00

  Yeah, Alpha Fold two was, you know, the biggest breakthrough, right? To be able to understand the structure of proteins. And then there was something called Alpha Fold two multimer, which then allows you to understand not just the structure of proteins by themselves, each individual protein, but the structure of proteins as they come together and what we call complexes, so how these proteins fit together. That opens up, you know, and helps us answer a lot of questions in biology. But there's still a big hop to designing therapeutics. And one of the, one of the big classes of therapeutics is what's called small molecules. So these are molecules that are not proteins. These would be things like caffeine or paracetamol, things that more often you can take as a pill.

• 27:24Speaker 02

  Hmm.

• 27:25Speaker 00

  And the way that these therapeutics work, these, these small molecules, is that they go through the body, they go into the cell, and they actually come and attach themselves to these proteins. You know, these proteins are the fundamental building blocks of life. They form these molecular machines by interacting with other proteins.

• 27:44Speaker 02

  Yeah.

• 27:44Speaker 00

  And so you can, you can imagine that if you have another molecule, your, your drug that comes in and attaches itself to a protein over here, then it might disrupt the ability for that protein to interact with another protein, all of its normal machine and day to day life. And so you're modulating the function of that protein.

• 28:02Speaker 02

  Yeah.

• 28:02Speaker 00

  With this small molecule.

• 28:03Speaker 02

  Yeah.

• 28:04Speaker 00

  And that's the essence of, of, of, of drug design and how therapeutics work. And so you can imagine as a chemist, your, your job, a drug designer, you're trying to design a small molecule that's going to fit to this protein over here and disrupt how it normally functions, or in some cases, enhance how it normally functions. And so it'd be really helpful to understand how this small molecule interacts with the protein.

• 28:30Speaker 02

  Yeah.

• 28:30Speaker 00

  What's the structure that it might make? What are the interactions, these literally physical interactions that are being made?

• 28:36Speaker 02

  Mm hmm.

• 28:39Speaker 00

  And so that really inspired, you know, the creation of Alpha Fold three, where now we have a model that not only predicts the structure of proteins, but how these proteins interact with small molecules.

• 28:51Speaker 02

  Yeah.

• 28:52Speaker 00

  Also other fundamental molecular machine building blocks, things like DNA and RNA. And this basically opens up the ability to structurally understand, which is a core part of drug design. Um, small molecules, it opens up new classes of targets. You know, there are things like transcription factors, which are proteins that sit on DNA. And read DNA, and you can imagine now trying to design a small molecule to change or disrupt the function of something like that. And so to do that, you'd really wanna be able to see, literally in three D, how this all looks.

• 29:28Speaker 02

  Yeah.

• 29:28Speaker 00

  And if I make changes to my little molecule.

• 29:31Speaker 02

  Yeah.

• 29:32Speaker 00

  How will that change the way it interacts with this protein in this biomolecular system? So Alpha Fold three is now very, very accurate. It allows us to answer a lot of these questions purely in silicone.

• 29:42Speaker 02

  Yeah.

• 29:43Speaker 00

  Purely on a computer where before you would have to go to the lab, literally crystallize this stuff. This can take six months, it can take years, sometimes it's even impossible. Now, at ISO, our drug designers are, you know, literally sitting, you know, with their laptop, browser-based interface, being able to understand, make changes to their designs and see the impact of that.

• 30:05Speaker 02

  Incredible. So there are a couple of interactions that Alpha Fold three is focused on. Proteins and nucleic acids, proteins and ligands, and antibody to antigen. Can you give us some good examples of, of the impact that Alpha Fold three now has on the interaction of these different types of proteins and molecules?

• 30:29Speaker 00

  Yeah, so protein and ligands, that's the same as protein and small molecules. So those two terms, ligands and small molecules, are synonymous. That allows us to understand how small molecule drugs interact. Then we can think about, you know, protein, protein interactions.

• 30:48Speaker 00

  There's a whole class of therapeutics called biologics. These are things like antibodies. That allows us to understand how they might interact with our targets. Opens up new modalities.

• 31:00Speaker 00

  And, and that also encapsulates the sort of prot- the antibody antigen interface. So if you're designing an antibody, you want to understand how your antibody design is going to interact with the protein surface there. So it's the same model that we can use across all of these different applications.

• 31:20Speaker 02

  What are the nuances of training a model like AlphaFold 3? And what are the benefits of using a diffusion-based architecture?

• 31:27Speaker 00

  Yes, a great question. There were a lot of challenges we had to overcome to get AlphaFold 3 to work. One of the most interesting things was actually just how do we

• 31:37Speaker 00

  take something like AlphaFold, which was only working with proteins, and then input these new modalities, these new data types of RNA, DNA, small molecules? So we had to work out how to tokenize not just proteins, which we kind of knew how to do, but how to tokenize then DNA, how to tokenize small molecules. For things like DNA and RNA, that's a little bit more obvious. We could tokenize in the bases. But then for small molecules, we would really go to, we tried a whole bunch of different stuff. It really ended up that this atomic resolution tokenization worked super well.

• 32:13Speaker 00

  And then you have the question of, okay, how do you actually

• 32:17Speaker 00

  predict the structure of this mixture of different molecule types?

• 32:21Speaker 02

  Yeah.

• 32:22Speaker 00

  You couldn't use the same framework as AlphaFold 2. And this is where diffusion modeling just really shunned.

• 32:33Speaker 00

  Here, we could actually model every single individual atom and the coordinate of every atom individually, and have a diffusion model be producing those 3D coordinates. And the tokenization that we talked about is conditioning the inference of that diffusion process.

• 32:53Speaker 02

  So interesting.

• 32:54Speaker 00

  And this was a huge breakthrough. So, you know, we're talking about on our leaderboard, just a massive step change, particularly in small molecule protein

• 33:06Speaker 00

  interaction accuracy. It was a massive step change and something that really unblocked the rest of the project.

• 33:11Speaker 02

  Wow.

• 33:13Speaker 02

  So data, compute, and algorithms. We know those three are important in all other adjacent fields. But I was surprised to read an interview with Demis where he shared that we're not data constrained in biology. Can you share your point of view on that?

• 33:28Speaker 00

  You know, I think it doesn't matter what field of machine learning you're in, you're going to feel some data constraint. And I think the point here from Demis is that it's it's not a real bottleneck. As in, we can make progress with the data that is out there, that the data we can generate. And real progress can be made. Um, it's not, you know, oh, we've got to sit and wait fifty years for like the world to generate data before we can actually make impact here. No, we're not seeing that at all. There are modeling spaces that, um, where the data has been sitting around for years.

• 34:05Speaker 02

  Yeah.

• 34:05Speaker 00

  That we that we can see that we can make, you know, really substantial progress beyond anything that people have experienced before.

• 34:12Speaker 02

  Yeah.

• 34:12Speaker 00

  Now, does that mean there's no opportunity for data in biology? Absolutely not. Like, this is going to be a fundamental part of of how we, um, you know, how we continue to develop these the these models and these systems will be. What data we go out and generate. Um, and there, I think there's just a massive opportunity. In my mind, um, the data, you know, the data for machine learning in biology hasn't actually been created yet.

• 34:41Speaker 02

  Yes.

• 34:42Speaker 00

  Yes, there's a lot of historical data, but there's a huge, but that historical data hasn't been created for the purposes of machine learning. And so when you're going out and thinking, how do I create data to actually train my model? You're thinking in a very different way to how people have gone out and generated data in the past. And that there's a big opportunity there to explore.

• 35:00Speaker 02

  What kind of data do you think we're missing here right now? And do we think, do you think that we need anything in synthetic data?

• 35:09Speaker 00

  Yes, so, um, I'm a massive fan of synthetic data. Actually, I have been for since the very beginning of my career where, you know, we would, I I was generating synthetic text data. Just to overcome the fact that, you know, I was a PhD student with access to a couple of thousand images and Google had millions and millions of images. And so instead, I just generated tons and tons of synthetic data and and that unblocked things. And, you know, we're seeing the same thing in the, especially the chemistry space.

• 35:39Speaker 02

  Yeah.

• 35:39Speaker 00

  Where we have. Good theory. We actually, you know, know a lot about physics. We know, you know, we have the theory of quantum chemistry and quantum mechanics and we can create simulators out of that. We can approximate that and create more scalable molecular dynamics simulations. This gives the basis for, you know, a whole host of synthetic data. Then we have the models themselves that, you know, especially we have generative models. This can actually generate data that, you know, we can use scoring systems to help,

• 36:12Speaker 00

  you know, really enhance the information content of this data.

• 36:15Speaker 02

  Yeah.

• 36:16Speaker 00

  But I think one of the the big open spaces will be on what's called in vivo data. So data that you would normally measure on a on a real animal, something like a mouse or a rat. We, you know, that you, there's some historical data on that, but you can't generate tons of that. You can't really generate any at all.

• 36:35Speaker 02

  Yeah.

• 36:36Speaker 00

  Right. So then there's a big opportunity to look to new data generating technologies. There are some incredible people doing things like organoids on a chip. So ways of starting to measure things that you would normally measure on a on a real animal, but, you know, completely on a on a chip. So, you know, I think

• 36:57Speaker 02

  So interesting.

• 36:58Speaker 00

  Yeah, there's going to be a whole host of like new breakthroughs in data generating technology in biology and chemistry. That's that's going to, you know, have big impact on how we think about modeling that world as well.

• 37:09Speaker 02

  Are you working on any of that internally or are you hoping that other players fill in some of that gap?

• 37:16Speaker 00

  So internally, we actually don't have any of our own labs in isomorphic.

• 37:21Speaker 02

  Yeah.

• 37:22Speaker 00

  Yeah, in isomorphic labs, but we we work with a whole bunch of other companies.

• 37:27Speaker 02

  Yeah.

• 37:28Speaker 00

  You know, we we generate a lot of data ourselves.

• 37:31Speaker 02

  Yeah.

• 37:31Speaker 00

  A lot of proprietary data. We've seen amazing impact of that.

• 37:34Speaker 02

  Makes a lot of sense. So there's a point of view that modeling structure of molecules and modeling their function and the modulation function is very important, but not necessarily always the limiting factor in drug development. What's your point of view on that?

• 37:51Speaker 00

  Yeah. As I touched on before, drug design is really, really complex. And as before, you even get to drug development, which is where you take those designs and you start putting them into real people.

• 38:02Speaker 02

  Yeah.

• 38:02Speaker 00

  Clinical trials. There are so many bottlenecks throughout this whole design and development space. Drug development is, you know, how how do we start to approach clinical trials? How should we test these drugs out in people?

• 38:19Speaker 02

  Yeah.

• 38:20Speaker 00

  How can we do this in a really timely manner, but still a really safe manner? Um, there's a lot of bottlenecks there.

• 38:26Speaker 02

  Yeah.

• 38:26Speaker 00

  That I think the industry as a whole, we will need to work out how to innovate in that space, especially as our predictive models of how these molecules will interact with people, how toxic they will be. As these predictive models get better and better, we will have to change the way that we approach clinical trials to really make use of that.

• 38:48Speaker 02

  Yeah.

• 38:49Speaker 00

  Ultimately, to get, you know, therapeutics into the hands of patients who really desperately need them.

• 38:53Speaker 02

  Yeah.

• 38:54Speaker 00

  You know, even in even in the design um of of of molecules themselves, as we talked about before, it's not just understanding the structure of these molecules.

• 39:04Speaker 02

  Yeah.

• 39:05Speaker 00

  It's not even just understanding how these molecules change the function of these proteins. But we need to understand how these molecules change the function of pretty much every single protein in our body.

• 39:16Speaker 02

  Right.

• 39:17Speaker 00

  Because if we take this as a pill, it's gonna go everywhere. And that's the major to- major cause of toxicity.

• 39:24Speaker 02

  Ah.

• 39:24Speaker 00

  Is when, yes, you you've designed this amazing molecule that like perfectly modulates your specific target that you know is key to your disease.

• 39:32Speaker 02

  But also affects other things.

• 39:34Speaker 00

  But it also affects other things. Um, now, of course, you do a lot of screening to protect against that. But um, the more we can predict that, the better. What's really exciting from my perspective is if we're creating these general models that understand how this molecule interacts with this target, but also any other target, then why can't we just use that same model to understand how these molecules interact with the rest of our body?

• 39:58Speaker 02

  Right, so interesting. So what is now possible with Alpha Fold 3 for drug designers? How are you using it internally?

• 40:07Speaker 00

  So Alpha Fold 3 gives our drug designers the ability to understand how their molecule designs really interact with this protein target. And this is the target of disease. And so our drug designers can make changes to the design. And then see instantly how that changes the way that this molecule physically interacts with the protein target. And that's really, really powerful. Before AlphaFold 3, you would be completely blind to this. You wouldn't actually probably know how your molecule is interacting with your protein. You'd be using your best intuition. Maybe somewhere down the line in the in the drug design project, you would get

• 40:49Speaker 00

  uh your structure crystallized with a particular design. That means going out to a real lab six months later, if you're lucky, getting a resolved 3D structure. Even then, that's just the 3D structure of a single design, not every single change that you make.

• 41:05Speaker 02

  Yeah.

• 41:05Speaker 00

  So AlphaFold 3 completely changes the way chemists can do this design work. But I was stressed that that that's that's nowhere near as far as we want to go. Because it's not just about what these molecules look like in terms of interacting. We actually want to know how strongly these molecules interact with this protein.

• 41:23Speaker 02

  Yeah.

• 41:24Speaker 00

  We want to know other properties of these molecules. We want to understand how the way that these molecules interact with this protein and how that changes the fold or the conformation of the protein, how that changes the function of the protein, how it might actually change the dynamics of the cell. There are so many questions, and these are these other AlphaFold-like breakthroughs that we're working on.

• 41:46Speaker 02

  Hmm.

• 41:47Speaker 00

  Um, that also go, you know, we have created incredible models for that our chemists are using in this design process.

• 41:54Speaker 02

  Interesting. So you're designing some drugs internally. What targets and programs are you focused on?

• 42:00Speaker 00

  So we have a really exciting uh internal program of of of drug design projects. These are focused on immunology and oncology. Uh, we've been making some incredible progress there. And it's been really exciting to see, especially how these models have transformed the way that we're actually approaching drug design on these these programs.

• 42:18Speaker 02

  You're also working with um Eli Lilly and Novartis.

• 42:22Speaker 00

  Yeah.

• 42:22Speaker 02

  And recently you announced an expansion with Novartis's partnership. Can you share a little bit about what these partnerships look like?

• 42:30Speaker 00

  Yeah, so we we signed these initial partnerships, um two partnerships, one with Eli Lilly, one with Novartis. That that was fantastic. They brought some really, really challenging problems to us. I think it's no secret that um, you know, the sort of targets um that, for example, Novartis brought to us, these are these are sort of targets that You know, the field and the Vartis, for example, have been working on for, you know, ten years plus.

• 42:57Speaker 02

  Wow.

• 42:58Speaker 00

  Um, so these aren't sort of, oh, we'll, we'll, we'll try things out problems. These are for real, you know, hard things. Um, last year was an amazing year, both for, both for our internal projects, but also for these partner projects to really see how well these models are working. It's allowed us to really, you know, uncover new chemical matter.

• 43:21Speaker 02

  Hmm.

• 43:22Speaker 00

  Um, working out new ways to modulate these, these targets that people have worked on for a long time. Um, you know, it's been amazing to see, uh, this new deal, which has expanded on the Vartis collaboration, which I think is a real testament to, um, you know, some of the success of, of the early days of these partnerships.

• 43:39Speaker 02

  Congratulations. I think it's incredible milestone, especially just one year in.

• 43:43Speaker 00

  Yeah.

• 43:44Speaker 02

  So I'd love to talk a little bit about the team. Um, you've built a truly excellent team composed of the highest caliber talent across many different fields. AI, chemistry, biology, and you've also brought in outsiders into the field to help question traditional thinking. Can you share a little bit about how you thought about this?

• 44:02Speaker 00

  Yeah, so the space of AI for drug design hasn't really existed for very long. Um, so the chances of finding a world expert at drug design, who's also a world expert in machine learning or deep learning is basically

• 44:21Speaker 02

  Zero.

• 44:22Speaker 00

  Zero, yeah. Just because these fields, these fields haven't coexisted, um, for long enough. I, I genuinely think about a new, um, sort of a field of science that ISO is breeding because we are, you know, we have these people who really live and breathe the intersection of this.

• 44:39Speaker 02

  Yeah.

• 44:39Speaker 00

  Um, so, you know, because, but because we can't hire these people, you know, I really think about how do we bring the world experts at drug design and medicinal chemistry and the world experts of machine learning and deep learning? And get these incredible people sitting side by side. Um, cause it's not just enough to have these amazing people sitting in their isolated teams.

• 45:04Speaker 02

  Yeah.

• 45:05Speaker 00

  We need people sitting side by side, speaking each other's languages.

• 45:09Speaker 02

  Yeah.

• 45:10Speaker 00

  With a lot of empathy, a lot of curiosity. Uh curiosity to understand this new science, to really build intuitions in your own language. And we've seen just such amazing things come out of this dynamic where you you really have, you know, a generalist machine learner who doesn't know anything about chemistry or biology.

• 45:33Speaker 02

  Yeah.

• 45:34Speaker 00

  Start to come in and understand the problems of a medicinal chemist and a drug designer.

• 45:41Speaker 00

  And um when I think about even hiring machine learners and machine learning scientists and engineers for the research that we're doing. I'd say, you know, sixty, seventy, eighty percent of the people on our team have no prior knowledge

• 45:56Speaker 02

  Wow.

• 45:56Speaker 00

  of chemistry or biology, maybe, you know, high school or university level. And that can actually be a real asset.

• 46:04Speaker 02

  Yeah.

• 46:04Speaker 00

  Because you come in sort of a little bit naive.

• 46:08Speaker 02

  Yeah.

• 46:08Speaker 00

  And as long as you're curious, I think one of the key things is asking, you know, the curious questions, asking this like stupid questions. And then and then that allows us to come at the problems from first principles.

• 46:20Speaker 02

  Yeah.

• 46:21Speaker 00

  It almost allows us to break through the the the the dogma of um, you know, previous experience and how people traditionally approach these problems. We can think ground up from scratch. Um and that's a lot of the mentality of how we think about creating these research breakthroughs.

• 46:39Speaker 02

  A little naive and highly curious and high agency is a very good thing.

• 46:42Speaker 00

  Yes, exactly, exactly.

• 46:44Speaker 02

  So in November last year, you also made a very big move in launching the AlphaFold server, um which releases code and model weights for academic use. Can you share a little bit about why?

• 46:55Speaker 00

  Yeah, so, I mean, AlphaFold has a long, long lineage of um being open for this, you know, academic and scientific use. And uh it was it was really important with this latest breakthrough of AlphaFold three that we make sure that this scientific community has access to this functionality because.

• 47:16Speaker 00

  You know, yes, AlphaFold three is going to be incredibly useful for drug design, it already is. But it's also useful for, you know, many other areas of fundamental biology and just understanding biology. And people are using these, people are using AlphaFold three server and modeling in very, very creative ways. Um, so, you know, it's very important for us to make sure that there is that sort of free use for non-commercial academic work. Um, and it's been incredible to see the, the, the, the take up of that and the use of the server.

• 47:48Speaker 02

  I'd love to talk a little bit about the future. Can you give us a tease of what else is to come with AlphaFold?

• 47:54Speaker 00

  You know, in, in terms of, you know, structure prediction as a problem, uh, I, you know, in, in my mind, I, I want to completely solve this. Uh, I think AlphaFold three is a fantastic step on the way of that, this significant breakthrough. Um, but, you know, the, it's not a hundred percent accuracy.

• 48:13Speaker 02

  Yeah.

• 48:14Speaker 00

  Um, what does even a hundred percent accuracy mean in this space? Like with a lot of areas of sciences, as you start to push the boundaries, you, you, you see that the problem opens up into even more problems. That's that, you know, that, that's the addictive part of doing science, right? Um, and I think that, you know, AlphaFold three is a good example of that, where as you start to get these capabilities, you see that actually there are even more deeper problems that we want to be working on, um, and, and stepping towards. So, yes, understanding structure better and better and more accurately is, is, is always going to be interesting for us. But then it's also, you know, not just necessarily about static structure.

• 48:54Speaker 00

  So AlphaFold three, it models these crystal structures, um, which are almost static crystallized versions of what these, these, these molecules, how these molecules interact. But in reality, we don't have crystals inside of us. We, you know, these molecules are in solution, they're moving about, they're dynamic. So you can think, okay, well, maybe understanding the dynamics of, of these systems is actually also going to be really interesting.

• 49:17Speaker 02

  Yeah.

• 49:17Speaker 00

  So, yeah.

• 49:18Speaker 02

  What does a G P T three moment look like in A I biology? And when do we get there?

• 49:23Speaker 00

  So if I think about G P T three. This is really a generative model, so something that's generating text. And the G P T three moment for me was, was, you know, crossing over that boundary between, yeah, we've got generative models of text and they generate some stuff and it looks like text, but I'm not convinced that it's generated by, by a human.

• 49:49Speaker 02

  Yeah.

• 49:49Speaker 00

  And G P T three started to be that first point where you're like, oh shit. This is like, this kind of looks like a human.

• 49:56Speaker 02

  Yeah.

• 49:56Speaker 00

  And so your, this generative model is actually recreating the distribution of data that it's trained on.

• 50:03Speaker 02

  Hmm.

• 50:04Speaker 00

  And what is a generative model? Generative model is something that fits the manifold of data that is trained on it. So when I think about this applied to biology,

• 50:15Speaker 00

  you can think about these generative models actually starting to recreate, you know, that G P T three moment, recreate what things would actually look like in reality.

• 50:27Speaker 00

  And that's quite exciting because

• 50:30Speaker 00

  that means that these models are spitting out things that either they actually exist in the world.

• 50:34Speaker 02

  Yeah.

• 50:35Speaker 00

  And we can kind of validate that, or maybe even discover new things that exist in the world. Or they could exist in the world.

• 50:43Speaker 02

  Yeah.

• 50:43Speaker 00

  Which means that they could be things that we could design or manufacture or create that would actually be stable and, and work and exist in our physical reality.

• 50:51Speaker 02

  Yeah.

• 50:52Speaker 00

  And I think the, the cool thing about this in biology is that unlike with language where with language, when it generates something at human level quality, we can understand that because it is human derived.

• 51:07Speaker 00

  But a lot of problems in chemistry and biology, we even struggle to understand ourselves. And so when we get to that G P T three moment, I think it will look a lot less like G P T three, but much more, feel a lot more like move thirty seven in AlphaGo.

• 51:19Speaker 02

  Interesting.

• 51:20Speaker 00

  Where we're starting to see things that are beyond human understanding, but that do exist in the real world, that exist in our physical reality.

• 51:31Speaker 00

  But are beyond sort of human comprehension.

• 51:33Speaker 02

  Right.

• 51:33Speaker 00

  And that's just going to be mind blowing. In fact, you know, we're starting to see that internally with our generative models that we're creating designs that a human drug designer would say.

• 51:46Speaker 00

  I'm not so sure about that. I much prefer this. And then you test it out in physical reality. And the generative model is correct and the human is wrong.

• 51:55Speaker 02

  That's fascinating. I love the move thirty seven analogy. When the model starts to see elements of creativity and

• 52:01Speaker 00

  Exactly.

• 52:02Speaker 02

  pass the human.

• 52:02Speaker 00

  Move thirty seven was this amazing move during the AlphaGo games against Lee Sedol. It was, you know, the thirty seventh move of the game and it stunned the world, stunned the Go world because it was uninterpretable by a human.

• 52:16Speaker 02

  Yeah.

• 52:16Speaker 00

  It looked like a mistake. No one had ever played this move in the entirety of, you know, thousands of years of human history playing Go. And it turned out as you unrolled the game that this was the critical move that allowed AlphaGo to beat Lee Sedol in that match.

• 52:30Speaker 02

  Yeah.

• 52:30Speaker 00

  And we're going to see so much of that sort of behavior coming out of these models.

• 52:34Speaker 02

  Yeah.

• 52:34Speaker 00

  Especially when we're applying them to things outside of of native human understanding like chemistry and biology.

• 52:41Speaker 02

  Yeah, I love that. Also our punchline today.

• 52:45Speaker 02

  So when will we see our first AI generated drug in clinic and also in phase one, two and three trials?

• 52:52Speaker 00

  So we're making amazing progress

• 52:55Speaker 00

  on our drug design programs and, you know, the thing I think about actually is as we start to get a whole bunch of these

• 53:07Speaker 00

  AI designed assets, these molecules, get into clinical phase. How can we actually start to think about engaging in that clinical development?

• 53:17Speaker 00

  To, you know, get these molecules to people, you know, as as as fast and as safely as possible. Because there's so much unmedical need.

• 53:28Speaker 00

  So, yeah, here I think about, you know, what are going to be new ways to engage with regulatory bodies? What are going to be new ways to incorporate our predictive models for not only how this molecule works for the disease, but how, as we talked about, how it interacts with the rest of the body? You know, the types of toxicity it may induce. I think there'll be a lot of opportunities to think about just. Uh, streamlining and speeding up this process.

• 53:55Speaker 02

  Yeah.

• 53:55Speaker 00

  Uh, maybe even completely changing the way we think about human clinical trials as we, you know, our AI models are become so, you know, we, we can design these molecules so much quicker in a much more targeted manner with so much more knowledge about how they work.

• 54:10Speaker 02

  Yeah.

• 54:11Speaker 00

  So that'll, that'll change the game, but I think we've got a long way to go as an industry to really work out how that changes.

• 54:16Speaker 02

  Yeah. Last question, as isomorphic succeeds and potentially as a whole field succeeds, what happens to the traditional world of pharma?

• 54:24Speaker 00

  I think they become, you know, in some sense, pharma will be using AI. I think, I think there's, there's no world where in five years' time, uh, you will be designing a drug without AI.

• 54:36Speaker 02

  Yeah.

• 54:36Speaker 00

  Like that is an inevitability. It'll be like, um, you know, trying to do science without using maths. AI will be this fundamental tool for biology and chemistry. It already is.

• 54:48Speaker 02

  Yeah.

• 54:49Speaker 00

  At least in, in isomorphic's world, um, that everyone will be using. So it's not going to be, oh, is it pharma or is it AI? That, it's going to be one and the same in the sense that the whole industry will, will adapt to that.

• 55:02Speaker 02

  Yeah. Amazing. Max, thank you so much for joining us today. This was a fascinating conversation.

• 55:07Speaker 00

  Yeah, it's been a pleasure. Thank you.