0:00Speaker A:Looks a lot less like a, you know, a ChatGPT and a lot more like Autodesk or SolidWorks or Figma. You know if you've used those things where you can kind of load up your molecule, there's this almost like Photoshop esque like design suite. You have this equivalent of paint tool to kind of paint your epitope. You have this equivalent of a content aware fill tool to kind of get your, your binders generated from Chai.
0:22Speaker A:And I think to add to that, right? Yeah. This notion of target discovery and hit discovery and optimization where each of these has a gate and takes a few months to a few years is this very like waterfall mod model where the cost of trying things and getting things early is very expensive. But I think to what Matt's saying, if you start to get in a regime where you can have models give you really promising candidates, you can start to make that look a lot more like a loop.
0:45Speaker A:Right. It's akin to becoming more agile in software development. But now the next problem is agonists. How do you reliably one shot hitting a switch on a cell.
0:54Speaker A:Right. Or by specifics or ADCs. Right. And I think this levels of abstraction that we're going to have to climb with the product as like the models get better.
1:04Speaker A:If you have like these really good primitives for structure prediction and binding and design and you can kind of compose them then you can start to just like grow into like the outer loop of science.
1:15Speaker B:Welcome to Latent Space AI for Science. I'm Brandon, I build RA Therapeutics at Atomic AI. I'm joined by my co host RJ Honacke, CTO and co founder of Mirror Omics. It's a pleasure to have with us in the studio today Mac McPartland and Neil Patil of Tri Discover.
1:30Speaker B:Chai's a protein design startup which is about 2 and a half years old and has made quite a splash in those few years. They have several very exciting announcements that I think they'll tell us about today. But yeah, to get started could you two give us a bit about your background and your what you do at chai?
1:46Speaker C:Yeah, thank you very much for having us. We're super excited to talk about CHAI today. I'm Matt McPartland, I'm one of the co founders of Chai. My background is in AI biology related stuff.
1:57Speaker C:During my PhD, I actually started my PhD in theoretical computer science and then trans transition to this later. Yeah, I, I've been doing this stuff now for like about eight years and I kind of came into the field at an interesting time where protein structure prediction was like just starting to see signs of life. So this is like AlphaFold1 days and was in the field during Alpha 2 and like got to see a lot of the interesting developments at that time. So yeah, I. I'd always been pretty interested in like applying this stuff in the real world and Chai was just a perfect opportunity to do that.
2:29Speaker A:And I'm Neil Patil, I help lead a platform and product here at Chai. So a lot of the stuff around infrastructure to train models, serve them, and then the productization piece, you know, the design suite that lets you use the models. I kind of have a more meandering path. So I kind of got into programming like 15 years ago making apps in the app store.
2:46Speaker A:Got really addicted to the dopamine hits you get from that and then actually got nerd sniped by robotics and like worked on that for a bit. Self driving cars in like 2018, 2019. Got really jaded and was like, I don't want to touch hardware for a while. Ended up switching and join a SaaS company called Vanta is one of the first employees there and kind of grew with it.
3:05Speaker A:Started my own security company afterwards. Got a few years into that and I was like, you know what, HADMs are kind of cool. Like I want to work on something a little more meaningful. And so I joined Chai about a year ago, right after Chai 2 was announced to help with a lot of the platform and commercialization pieces.
3:21Speaker D:Awesome. It's like the five stages of grief or something.
3:24Speaker A:Yeah, yeah, we're at acceptance.
3:27Speaker D:Awesome. You have these, I think four now, big partnerships and raised a whole bunch of money. Can you tell us a little bit about those partnerships? And then what I really want to know is what are you telling investors and customers that is so compelling that they're willing to do these big deals?
3:45Speaker C:Yeah, so like we've been very fortunate to partner first with Eli Lilly and then with Pfizer, Novartis and our Gen X. Yeah, I think it's been like a really interesting ride. And I think our business model is also very compelling to a lot of people. Like, we really like to, we care about the partners succeeding like this. Chai as a company really depends on how the partners succeed.
4:07Speaker C:I think Neil probably has some interesting takes on like, you know, what we actually offer and what makes that so compelling. So I'll hand it over to you.
4:13Speaker A:Yeah, I mean, as you all know, drug discovery is a very lengthy process right And a lot of these pharma companies are spending lots of time, you know, years and years and billions of dollars trying to find initial therapeutic candidates. And, and so at Chai, you know, we train models that can help accelerate that process and kind of find those initial binders and then some. And you know, we, you know, there's a lot of bio companies, AI for bio companies that are like making their own drugs. We really don't see ourselves that way.
4:38Speaker A:Right. We see ourselves as almost a neutral software factory for making medicines. And so that's what, you know, lets us go then work with and support all of these other farmers in their kind of drug discovery journey. And so, yeah, I mean, a lot of this capital is just another proof point that we can sort of start to really accelerate that software factory, right?
4:59Speaker A:Go after harder modalities, train bigger models, and ultimately just build what our partners and customers ask us for.
5:05Speaker D:But what is it that why you and not other structural companies? Why are, why are they compelled to buy from you?
5:13Speaker C:The thesis of Chai has always been to like, be the software and modeling layer, which was, I think, like, very controversial at the time. Like everyone, you know, this, this play.
5:22Speaker B:Is definitely two years ago and it's already like a completely different world.
5:25Speaker C:Yeah, yeah, it's, it's pretty crazy. Like the, like, people tried this play for a while and I think like, the models just really weren't there yet. And even like, for us, we were taking a risk in the very beginning. Like, we were kind of banking on the models getting there.
5:37Speaker C:And like, I had seen early signs of life in my work and our CEO Josh, like, he, he was on the original ESM papers on that team at Meta, and he was seeing like pretty early signs of life that like, you know, there might be scaling laws here. They like, I think we'll actually be able to start like designing things. Structured prediction is getting really good. Like one, one like crazy thought is like, we didn't have a multimer structure prediction model until like 2021, that's five years ago when we could like start with deep learning to like actually predict the shape of two proteins at once.
6:06Speaker C:Like it was a fold one was like an alpha two was like this huge breakthrough. But then like Alfal 2 multimer came out like a year later. So like, you really kind of needed that to unlock design in the first place anyway. Like, we weren't even trying to predict multiple proteins at once.
6:20Speaker C:And then really like around that time inverse folding kind of started working and it was like, oh, protein mpnn this actually works in the lab. Like credit to the Baker lab for doing all of this really excellent lab validation and all their models. But I think like, we're starting to see them do interesting things and like actually work on like real world experiments. Now is probably the time to start betting on this.
6:40Speaker C:I think like before then maybe you could take like some experimental data from a campaign on like this one target that you had and you care about and you, you might be able to like make some progress on that and like keep hill climbing in this like one very specific case. General models weren't really a thing back then. So I think like, yeah, we took that bet pretty seriously and like we, we decided to just like push as hard as possible and to really like shoot for generality in our approach. And then when Chai 2 came out, our second paper after Chai 1, we kind of like show the world like, this is actually possible and it's possible at scale.
7:14Speaker C:We didn't show this for like one or two targets. Like, it kind of works. Like we were like, let's just go all in. I think Josh likes to say we set up bold company wide challenge to design antibodies to 50 targets.
7:25Speaker C:And actually like, we saw some signs of life. We're like, all right, let's like, let's do this with real statistics and see if this actually works. It's an interesting story of how we chose these targets. So we were like, all right, what targets we're going to choose?
7:36Speaker C:We should choose like some interesting targets, whatever. And at that point we're like kind of ramping up with CROs and figuring out like, what, what does our wet lab process look like? And we decided after, after trying some stuff, like many proteins, whatever, we're like, here are the interesting targets, this is what we should look at. And like half the time the targets just like kind of didn't work.
7:55Speaker C:We were still learning, whatever, and we're like, all right, maybe we should just go with like targets that the CROs have actually validated. So let's get the CRO catalog, see what they've already worked to unrestrict that to like an interesting set. So from that we chose 50 targets, designed antibodies against them, got hits to half, and at that point I think pharma started to realize like, okay, there are actually signs of life here. And this, this might actually work in some of our programs.
8:18Speaker D:And so antibodies is maybe a more challenging domain than other structural prediction problems. But so why tackle antibodies? So maybe back up, what is an antibody?
8:31Speaker D:And what do you do with it that and why is it an attractive target?
8:36Speaker C:The analogy that everyone gives like this lock and key kind of problem where like your target, this protein that you're trying to bind to it might be some like disease protein that's kind of like your lock. And then you want to design this key that fits into it and like in our case just like sticks there. The interesting thing with the antibodies is like these like really flexible general proteins, like in a lot of ways they're very general. In a lot of ways they're actually like pretty uniform.
8:59Speaker C:But at least like how they bind to a target is very general. So like you have a lot of optionality in how you design this kind of binding interface. The structure prediction problem for antibodies, like predict how this antibody actually binds to the target, how it, how the key fits into the lock. That's been a notoriously difficult problem.
9:16Speaker C:The nice thing is like, so we've made a lot of progress on structure prediction. Kind of the field as a whole has come a long way along like in getting structured prediction to where it is. But in the design setting you can be a lot more selective about the types of designs you want to make and the types of structures you actually want to focus on. And in some cases it might actually be even easier to design a protein binder that is an antibody than to actually predict how it might bind that target in general.
9:44Speaker C:So like it's kind of like if you, if you have the freedom to choose, you can kind of just pick the easy cases. That makes sense.
9:50Speaker D:So the antibody is like there's a whole machinery in the body that works with antibodies. What, what does the body do with it naturally and what can you do with them? That is sort of not natural but is useful for therapeutics.
10:04Speaker C:This is coming from, from a non biologist here, but I think anybody's like they're these kind of like Y shaped proteins, like kind of looks like a peace sign with your fingers. Each, each of these fingers is kind of like uh, an arm of the antibody. And like it's really actually only the tips of the, of your fingers, the tips of the antibody that engage in binding. So this makes these really like nice therapeutic design targets for that particular region.
10:27Speaker C:Reason the nice part is that like the rest apart from the tips is like actually relatively constant. So this is called like the framework region of an antibody. And the design problem you're typically just designing like the very fingertips and you can actually choose for the most part like these kind of framework regions that your immune system already recognizes. So antibodies kind of like these Y shaped proteins that your immune system like recognizes, it knows really well.
10:52Speaker C:It's kind of like your body's. It's one of the lines in defense against pathogens and other types of diseases.
10:58Speaker D:So. So I guess antibodies can on the one end like connects to proteins on the surface of a cell typically or other things, but typically on the surface of a cell and then the other end helps the immune system identify a pathogen typically. But you can also do things like you mentioned ADC's anti antibody drug conjugates. So that means putting a drug on the other side or something like that.
11:23Speaker D:And that causes the. When you bind to something that it releases the drug into the cells, right.
11:28Speaker A:They're like this very general framework, right. Where kind of on the ends you have these CDR loops and you can design them to kind of bind to arbitrary things where maybe one end you bind to a cancer cell, the other end you bind to a toxic molecule. You're now precision delivering that toxic molecule to a cancer cell. Right?
11:45Speaker A:Or you just have two ends bind to things and kind of force like induced proximity to have some effect in the body or you know a lot of drugs historically are really just like about like blocking things. Right? Like anti agnist behavior. Right.
11:58Speaker A:But maybe you can have agonist behavior. We actually like really precisely like press a switch, like there's a GPCR protein which are these like doorbell proteins that sit in your cel brain. You have an antibody like very precisely engineered to poke it in a certain way that causes a downstream chain reaction. And I think one of the things that's really exciting about where we're getting to with some of these models is we can start to get that precise.
12:22Speaker A:Right. We can really target a very specific epitope meaning like binding spot. Right. A very specific set of atoms to have the antibody go after.
12:30Speaker A:Which historically with a lot of drugs you're just kind of brute forcing a lot of antibodies and just trying to come up with a bunch of things and see what sticks. But maybe that gets you a binder to some spot of your target molecule. But that doesn't let you precisely engineer where you're poking after.
12:45Speaker B:I know you're not biologist, but do you have any like idea about how they used to design these before, you know, these models came up? Like what would you. What was the grueling process you would do?
12:56Speaker A:What is the grueling process?
12:57Speaker C:Or what is.
12:57Speaker B:Which is actually still. Yeah, what still is the state of the art in terms of drugs which have made it to the Clinic.
13:03Speaker C:Josh, our CEO, likes to say that our, our biggest competitor is the mouse. So like our nature certain ways. So like traditionally these, these types of like drug like molecules were either discovered in like these immunization campaigns. So like, you literally will just like infect a mouse with a disease and see what antibodies it makes to try to like combat that.
13:23Speaker C:Other ways of doing this is like super large yeast display, so on. So you might like start with, hey, I really like this framework and how am I going to like figure out the right loops to design to bind this target? I'm just going to try as much as I possibly can and just like literally search for a needle in a haystack. And this would be like on the order of like at least billions of potential molecules that you're screening against this one target.
13:45Speaker C:And in that case you might like end up with, you know, one, two, maybe like a dozen potential hits to this target. You actually, you don't know much about this hits. All you know is that they kind of like stick to the target. You don't know necessarily where or like if they're even necessarily drug like.
14:00Speaker C:I think like one big separator of chai and like a thing that definitely our partners like to see is like, you can be really intentional with how you want to do this, this design process. You can say, I want to bind this target in this particular area. You can even go back and look to the designs after, like we validated that our designs. So you can go back and look and say, like, is this antibody engaging the target in the way that I expect do?
14:20Speaker C:I think this will actually have the therapeutic effect that I'm going after.
14:24Speaker B:One of the cool things about knowing that you have the right binding pose is that you can now also design selectivity into that. Does your platform have some technique for doing selectivity?
14:35Speaker C:Yeah, there's a nice mix of ideas that went both into the modeling side and especially on the product side for dealing with selectivity and cross reactivity. So in some cases you want your molecule to bind one target and avoid another one. So you might have healthy variants of protein and disease variant of protein. You want to avoid this disease variant, or you might have some other similar protein that's not actually harmful in your body that you don't want to just like artificially block.
15:03Speaker C:So I think like on the modeling side, yeah, we've come up with ways of doing that, but I think it's even more interesting on the product side. So like, how do you enable customers go through or partners to Go through and like, actually intentionally design for these things.
15:15Speaker A:Yeah. And maybe to like, back up and define cross reactivity.
15:18Speaker A:Like, it turns out when you're developing a drug, you're not necessarily going straight to injecting that into a human. Right. Like, you might want to put it in monkeys first, for example, and the monkey might have a maybe mostly similar, but slightly different variant of it. And so your drug not only needs to bind to the human variant, but also the monkey variant.
15:36Speaker A:Right. And so, you know, the way we've tried to model the models and the product is to kind of let you account for those very general cases where you say, hey, I'm trying to design something that can bind to both of these things so that I can actually go and develop the drug. Let me actually identify maybe the region that's conserved and then target conserved means, you know, doesn't change much between the two and target that exact region and then, you know, similarly with. Across with selectivity.
16:00Speaker A:Right. Maybe you might want to. There's a very similar protein in the human that if you accidentally bind that one, that's very bad. And you only want to bind the target protein.
16:09Speaker A:And, you know, that's why a lot of drugs. Right. You know, fail or are toxic or have, you know, really bad side effects. Right.
16:15Speaker A:And so it's kind of. You're kind of having this like, combinatorial problem of like, you know, bind only these things and avoid only these. And I think what's been really exciting with some of the progress recently has been, like, a lot of the improvements we've been able to make on the level of specificity we get, we can get to with those models.
16:33Speaker D:So you're not only designing the bind here, but you're also making sure that it doesn't bind to another.
16:39Speaker D:Other ways, like Car Ts have tried to tackle this by having some molecular or some sort of signaling pathway that says if I bind, I only fire if I bind. This one binds and this one doesn't bind. But you're saying you just design a, an antibody that actually only will bind to the thing that you care about.
16:58Speaker A:We're getting to the point where in some case, I mean, it's nuanced. Right. But in some cases you can actually try that.
17:02Speaker D:Okay, that's amazing.
17:03Speaker B:Yeah. So. So you, you're saying you essentially call it counter screen or you have, in part of your platform, you can now reliably counter screen against like a large, diverse set of proteins, which might be issues for downstream.
17:16Speaker A:Yeah, I would say the Framing is more. You can be very specific about what you care about binding versus what you care about avoiding. But I think, you know, for example, like, a lot of the money that we're raising now will let us train bigger models that can maybe be even more general and start to account for even more things at the same time. Right.
17:32Speaker D:Maybe we should back up. Let's talk about, so the history of the CHAI series of models. Well, why don't you tell the story?
17:41Speaker C:We started CHAI around two and a half years ago. The first couple months we're like, all right, we're going to work on protein design and we're working on this. We're making some progress. We're like, oh, it's pretty interesting.
17:52Speaker C:Like we had some ideas and models and then kind of like that was right when Alpha 3 came out and we were, we'd like been talking about like, man, we really need like an MSA pipeline. We need like all of this infrastructure. MSA is multiple, multiple sequence alignment.
18:06Speaker D:Why is this just. We've covered this before, but what is a msa? Like in two sentences and why is it important?
18:11Speaker C:So if you want to predict the structure of a protein, it might be really useful to see a bunch of very similar protein sequences. And what those protein sequences that are really similar tell you is like kind of what positions, like which amino acids end up being conserved across many variants of this protein. And if you see like high levels of conservation or like kind of high levels of mutation, like correlated mutations, that typically gives you some indication that these amino acids are close in 3D space. So you kind of have this like 2D view of a protein which can then be used to help you predict this 3D structure.
18:42Speaker D:So you're learning from evolution what was conserved, because the things that weren't conserved probably broke the protein and something died or didn't make it.
18:50Speaker C:Exactly. Right. Yeah, yeah, it's, it's, it's pretty remarkable that this works, honestly. Yeah, one of my favorite, like, biofacts here.
18:57Speaker C:Yeah. So, so we were like, kind of thinking like, oh man, it'd be, it'd be nice to have like a lot of infra and whatever. So Alpha 3 came out. We're like, hey, we should, we should like open source this model.
19:06Speaker C:We should just like, you know, bunker down, build all the infra that we need. I think like, this will pay back, like in the long term for sure of just like, as a forcing function to like, be where we are and also just like to contribute to the community as a whole.
19:19Speaker D:So it's interesting that you chose. Okay, this, we're actually, what we're doing here, we're building a model, but what we're really doing is learning how to build the infrastructure. Is that kind of what you're saying?
19:27Speaker C:Yeah, that's exactly right. And, like, I, I had built a lot of, like, similar infrastructure in my PhD, but not at a production level for a company. So, like, at that point, I think we were five people. So there are five of us at chai, and we're like, all right, this is our forcing function.
19:42Speaker C:We have, like, a clear goal to work towards. It's like, very direct. Let's get this thing going and see how fast we can do it.
19:47Speaker D:You guys were at this time sitting in the open AI offices?
19:50Speaker C:We were sitting in the open AI offices, yeah. In the, in the mission.
19:52Speaker D:Right. So, like, what's the backstory on that is really interesting.
19:55Speaker C:Two of our other co founders, Josh and Jack, had a relationship with some of the OpenAI people. Actually, OpenAI Co led our seed round. So we were like, kind of thinking, like, all right, should we get an office while we're only five people? And it turned out like that office was mostly vacant.
20:10Speaker C:So we, we got to sit in on the, like, in the open A offices for a while.
20:14Speaker D:CHAI one built it open source, learned about infrastructure.
20:19Speaker C:So then after that, like, we, we really set the sights down on protein design.
20:23Speaker A:And worth pointing out, Chai 1 was a structure prediction model. Right. So you have the, you have the sequence. What is the structure that it folds to?
20:30Speaker A:And then that was.
20:32Speaker C:Yeah, CHAI one. CHAI one's finished one. One other crazy story there. Let's see if we can actually share this.
20:38Speaker C:But this is a hilarious one. So, like, we, we were like, oh, man, we really want to be the first to put this out. And we were like, okay, we're. We're one week out.
20:46Speaker C:We're like, the model is, like, almost done training. We're like, should we. Should we build a web server? And then we're like, oh, yeah, maybe not.
20:52Speaker C:And then, like, we ended up spinning up, like, this whole web server so, like, people can use it, like, rather than just, like, download the git repo. It's kind of annoying, especially for biologists. And, like, we actually wanted people to use this, so, like, let's spin up a web server. Let's get the technical report out, all this stuff.
21:06Speaker C:So we, we ended up like, we were up for like, 48 hours straight, just, like, getting the paper over the line, getting the, like, all the last things done on the web server. And then Josh was interviewing with like Bloomberg TV or something that morning and we've been out for like 48 hours straight. So Josh like runs into a room to do this interview on Bloomberg tv. And like I think it was like seven in the morning, everyone's in the office.
21:28Speaker C:Like we, we didn't like want to be seen, whatever. And like the interviewer is like, oh, like interesting company. Doesn't look like there are any employees here. But yeah, it was, it was a really fun time.
21:39Speaker C:I think like the early startup days were just, just super fun. So yeah, after that we kind of set our sights on design and really what we were thinking is like we kind of always had antibodies in mind. We thought of this as like the most tractable problem. The nice thing with proteins is you, you have this beautiful sequence representation.
21:56Speaker C:There's already a lot of research been done in like autoregressively generate sequences. How do you like this Sequence generation problem is well studied. So we were thinking like what, what's a nice like area to apply sequence generation to in, in the bio space? And it, it's pretty natural to do like linear sequences of amino acids.
22:14Speaker C:So we start working on design. A unique thing about CHAI is like we're not like we're designing antibodies. Like we're an antibody company. Like we don't, we don't really like pigeonhole ourselves into like one therapeutic area.
22:25Speaker C:So we like tried to really tackle this problem very generally. So we were thinking like, can we design many proteins, can we design antibodies, can we scaffold regular complexes? So like really just take a holistic view on like how do you design proteins in general? And that eventually led to the Chi 2 model.
22:42Speaker C:So that was our like first flagship design model. And that's where the Chai 2 paper and like our Bold Target discovery project came in. So we design antibodies to 50 targets. For that paper got binders to about half of them with I think on average around a 20% hit rate for binding.
23:00Speaker C:And then afterwards started working on Chi 3. So that's our latest series of model, but I'll break there.
23:06Speaker D:But before we talk about Chi3, can you tell us about, especially for listeners that may not be familiar with structure prediction models, what does the model look like? How does it work in general?
23:15Speaker C:Let's take a look at Chi 1. Chai 1 has this like roughly a tokenizer, a transformer, something that looks like a language model and then something that kind of looks like an image diffusion model. And they're all just like stick together, stitched together. The tokenizer is like not your kind of typical like words of X style tokenizer.
23:33Speaker C:This is like I have a bunch of atoms in a molecule and now I want to like pull those into what I would call tokens for my like LLM looking trunk. And then that conditions this like kind of big diffusion model which will then emit the image which is some 3D structure.
23:48Speaker D:So is it atoms or is it amino acids that are the input?
23:52Speaker C:It's an interesting question as well. So we have like all these different input tracks. So like one thing about biology is the data is inherently multimodality in a sense. You have these, this like, you know, kind of token sequence representation.
24:05Speaker C:Each of these tokens has like a set of atoms that kind of dangles off. And then you also have you know, some, some properties of the different atoms. Like an atom is, might have like a different charge, it might have a different element type. So like periodic table of atoms.
24:20Speaker C:And then these kind of all get bunched together into tokens. Once tokenized you can kind of process this in very standard ways. But then ultimately you have to get back to these like 3D coordinates. So like in order to predict the structure, this is just some 3D object and that object goes through or like to emit that object you go through what looks like an image diffusion model where you kind of go back from, from tokens back to the atom representation.
24:44Speaker D:I see. So the tokens go in, the transformer establishes the relationship between the different tokens and then the diffusion model turns that represent that latent representation into a 3D structure.
25:01Speaker C:That's exactly right.
25:02Speaker D:Yeah. Okay, great. So that's Chai 2, that was Chi 1. Okay, Chai 1.
25:06Speaker C:So Chai 1 folding model. Yeah, it's like uh, and like all this bio stuff, it sounds like kind of scary or like atoms, tokens, amino acids, like at the end of the day, my background personally is like theoretical computer science. That's what I spent like all of my earlier years doing. Transition to this like pretty late in my PhD.
25:24Speaker C:But I think like the background that you need is really similar to the background that you'd need for like any other field of machine learning. There are all these domain specific things that you learn about, but like one, one analogy or like anecdote I like to like to say is people think you can't work on like AI bio unless you're a biologist. But it's kind of like you can't work on like video models unless you're like a director or something. Like there are all these like super domain specific things like oh yeah, to understand like lighting and a video, things like that.
25:50Speaker C:But at the end of the day these are just machine learning problems and like they're, they're all solved the same way.
25:55Speaker D:Okay, so then Chai 2 there's a jump in capability as well as an architectural change, right?
26:00Speaker C:Yeah. What we've disclosed about Chaito is like it is an all atom diffusion model. So we, we're trying to predict like you know, atoms in 3D space still. But we're doing it in such a way that like the model actually has the ability to like design atoms, place them, decide which atoms actually are there.
26:17Speaker C:So like one way to represent amino, an amino acid, like a protein token is by like which atoms are present. So in the Chi2 case we were just predicting like all right, show the model. Let the model just kind of pick what atoms it wants to keep and then map that back to what amino acids there are.
26:34Speaker D:What are you able to do with Chi 2 that you can't do with Chai 1 is just like better or is it, are there new capabilities?
26:41Speaker A:It brings it's design, right? So Chi1 lets you say, hey, I know the sequence of amino acids, right, that text string and I know the.
26:49Speaker D:Structure that you would get from like the genome or.
26:51Speaker C:Right, exactly.
26:52Speaker A:Chi 2 says okay, I have a target structure that I want to design, a binder to try to then generate candidate molecules, candidate medicines that bind to that target. And so this is kind of a design model or design family of models. And I think that's where you really cross the threshold of usefulness, right? I mean Chi1, Alphavol very useful because you can at least intuit and reason about the structure and see what you're looking at.
27:17Speaker A:But the ultimate goal here is to design medicines, right, and design new molecules. And I think Chi 2 really crossed the threshold of performance for doing that with antibodies a year ago.
27:27Speaker C:One analogy here would be like kind of like back to the image domain. So like Chi one would be like, you know, there is a cat in this image. Like thanks Chai 1 and Chai 2 is like I'll show you a background, maybe like I'll prompt you with some, some like image information. Like hey, put a cat in a field and Chai2 will actually just like give you back an image of a cat in a field.
27:48Speaker C:And you're like that that's a good looking image. Or it's not. You might have some other model which kind of ranks the image, but fundamentally it's the generative problem.
27:56Speaker B:So there's. Taking an analogy a step further, it's maybe more like you showed a background and then it generates, there is a cat and then it generates an image of the cat at the same time. And it makes sense that there is a cat in this field and also that the cat works in the image. So there's a, it is a, it's an interesting problem because you have to generate two things at the same time, both the sequence and the structure.
28:17Speaker B:Can you, if you, I don't know if you can, but could you talk a bit about like how that works? Like how do you do that? So you code, you co design the sequence in a way that the structure also fits and makes sense.
28:27Speaker C:One way to think about it is kind of like the classic way of doing this. Let's talk about both in structure prediction. Like, all right, I know the sequence and like I can from that roughly figure out the 3D shape. And then there's kind of like the inverse folding problem which is like given a 3D shape, give me back a sequence that would fold into this.
28:44Speaker C:And now you kind of like need to do both things at the same time. But I think like similar principles apply. Like you can kind of have the model like think a little bit about what should this structure look like. Then you can have some other part of the model thinking about like now what sequence would maybe support this?
28:57Speaker C:And then like a nice thing with diffusion is like you can do this pretty slowly and pretty iteratively so you can give the model a lot of time to think about, all right, if I change the structure like this, how should the sequence change? And you can kind of just play this back and forth and back and forth and eventually it ends up kind of converging on something that's self consistent.
29:14Speaker D:It's almost like an EM algorithm.
29:16Speaker C:Yeah, exactly.
29:18Speaker B:So you have this model now, Chaitu, which is able to predict or to, to sample a structure and a sequence which generates that structure. And just because you can generate a structure like, doesn't necessarily mean it's necessarily accurate enough to do something. So do you have other scaffolding on top of that? Are there additional problems like are you one shotting these things or are you, you know, needing to generate thousands of them?
29:44Speaker B:And then you have a ranking or scoring or you know how like just having a candidate is maybe, let's say not enough? Um, so what do you do once you sample a structure or traditionally what's, what's Done.
29:56Speaker C:And like when, when co design and like protein structure design like started to become a thing, we're like kind of at a loss for metrics. Is like how do you know that your protein like you designed some, some like sequencing structure? Like how do I know that this is legit or not? Like I can tell you it's like anything.
30:12Speaker B:It's like totally out of domain now, right? Almost by definition.
30:15Speaker C:Yeah. And like as a human you can look at this thing and be like, I, I don't know, it checks out. Like even biologists are like, I have no idea if this thing actually folds, like maybe some of it. Even our biologists are surprised by the way, with some of our designs that do end up working.
30:28Speaker C:What was done at the time is we kind of came up with a bunch of metrics and alphafold really is what enabled this. So you'd take the sequence that you predicted, you'd run that through some totally distinct structured prediction method. So this is completely independent of your model. And you say if an independent model thinks that this sequence folds to a similar structure, then it has a higher likelihood of being correct than just whatever the prior likelihood would be.
30:54Speaker C:So you can take your sequence now and you can measure like how consistent is this structured prediction method with the structure that you actually predicted for that sequence. You can now compare your design to an independent model structure prediction. And that became like a really good way of gaining conviction that your, your design model was correct. And people kind of like game these benchmarks for a while and kept pushing, pushing, pushing.
31:16Speaker C:It turns out like it's easy to get self consistency consistent design to structures. If all of your proteins look identical there, there are a lot of problems that this, this creates. But then people started adding more and more on top of this.
31:27Speaker B:Yeah, that is a, that's an interesting point that I think some people have acknowledged in the community. So how did you solve that? Yeah, you can see that if you sort of use your oracle and also your sampler at the same time, you eventually will converge. What do you do to stop that or to convince yourselves that you're doing something valuable?
31:45Speaker C:One of the nice things about structure prediction methods is that usually you have some calibration how kind of how confident the model is in its prediction. It turns out these models, they can give you a pretty well calibrated confidence prediction. So rather than just say this is what I think the structure looks like, it'll say this is what I think the structure looks like. And kind of like here are the parts that I'm not really certain about.
32:08Speaker C:And you can kind of aggregate this down to like a single scalar. And typically what people do is they'll look at like okay, like not only how self consistent am I, how much does this independent folding model even like the structure that it outputs? Uh, so that was one way of early on I, I'd say to like just gain confidence. And then like another thing that people often do is they'll look at like the diversity of their generations.
32:28Speaker C:Because again you could have a model that's perfectly consistent, gives you great confidence. Predictions back might be the same structure every time, like same sequence every time. So you also want to see like, okay, how diverse are the solutions? How many of these new problems can I solve in a sense if I.
32:43Speaker D:Had a lot of, a whole lot of money to validate, how would you do that? Can I go and you know, do cryo em or something like that and try to figure out the structure, you know, sort of get some ground truth on that.
32:56Speaker C:It's more that the feedback loop is really slow. So you can validate a few structures like this, but it might take months and it's, it's just not like a very scalable direction. So I think that's like a problem for the field as a whole. And I think people are spending a lot of time even like especially at chai, I think thinking about how do we validate these, these problems at like bigger scale, how do we you know, basically increase the throughput of our validation or increase the PS cycle time.
33:18Speaker C:Because if you're waiting months to figure out, hey, was my model correct? Like it's just, it's hard to iterate in a research environment that way.
33:24Speaker A:The good news is that this is getting a lot better, right? Like there's a whole network now of wet labs that you can work with that will run, you know, these assays, these experiments and tell you things about say, you know, does your protein that you came up with bind to its target? Well and so you know, thankfully we're not at years, right? We're down to like weeks, which you know, not as fast as like LLM land where you can just you know, scale up in eval and throw more compute and get results back in hours, but you know, fast enough to where you can start to recursively self improve.
33:52Speaker A:And you know, I think we also spend a lot of time like you know, figuring out what are the metrics that we can compute, you know, in silico like on the computer that are predictive perhaps of lab success. But you Know your question about cryo em? Yeah. I mean, also, you kind of have to measure the structure.
34:06Speaker A:And as you know, that's like so expensive because you have to kind of freeze the protein and shoot these electron beams at it and see how they bounce off. I remember there's like this really funny anecdote. We'll see if I can share it. But like the, you know, the paper in Chi 2, we actually, you know, did that, we took some of the, you know, the, the proteins that the, the model predicted and, and ran cryo em and we got the results back and we're like, wait, the, the results look wrong because we, we had overlaid the kind of prediction over the point.
34:33Speaker A:The, the, the electron cloud, the point cloud. We didn't see any difference. And point being like, we're, we're getting the point now where these structure prediction models are within, you know, a few angstroms or less of the actual atomic positions that you validate.
34:45Speaker C:And in this case, it was a 0.33 Angstrom error, which is one third the width of an atom. Yeah. And we were like this, this, like, can't even be right. Like, clearly they just sent us back the wrong design.
34:55Speaker D:They just sent us back our design.
34:57Speaker C:Yeah, exactly.
34:58Speaker B:Did you check for data leakage?
35:00Speaker C:Yeah, in this case, like, there were no. So like, we actually chose these targets specifically, like to have no known antibody binder. So like, if we did get a hit, like, it was definitely the first antibody hit to this target. Yeah, I think that's one of the.
35:13Speaker A:Things I didn't realize about biology was like, just how much of it is literally feeling around in the dark. And that's not even a metaphor. You literally can't see, like, how these things look. Right.
35:23Speaker A:So structure models are so, so huge because now you can, okay, you can actually predict within an atom, you know, how these things look. And that enables you to then do things like Chai 2 with the design models.
35:33Speaker D:This, to me is AI for science is one of the cornerstone problems, right. Is that you don't know, you fundamentally don't even know how to measure your problem in a lot of cases. So it's very difficult to validate.
35:47Speaker D:So you, you, you're getting these sub angstrom predictions with Chai 2. Chai 3. What, why Chai 3? What's better or what?
35:55Speaker C:Yeah, I think like with, with Chai 3. So like, honestly, like, there was a Chai 2, there's a Chai 2 and a half, there was a Chai 2.7, there was a eventually a Chai 3 and like each time we saw better and better performance. And I think like the, the main thing with Chi 3 is like we look at Chai 2 and like we look at the targets it could solve. There was like a lot of internal discussion after Chai to you, like, hey, we made like successful molecules, binders to half of these 50 targets.
36:21Speaker C:What about the other 25? You know, what can we do to make those better? And then like, you know, we were split. We're like, all right, should we like study these targets that we miss and like figure out exactly like are there properties of these that we can look at or should we just bet on the models?
36:34Speaker C:Like will the models just get there if we put more time into like, you know, just be bitter, less impaled in that sense and just really bet on the models getting better? And we, we definitely took the the latter approach. Like we bet on the model is getting better and we just pushed as hard as we could on that front.
36:48Speaker D:So you're scaling up the model, the data, whatever to, to just build more accurate models.
36:56Speaker D:Is the accuracy, is that the main thing? Is it binding affinity?
37:00Speaker C:What, what do we, So I think binding affinity is a big one. Like you, you, you, you can't just bind weekly. In order for this to be like useful tool, especially for our partners, we need to start producing molecules that are like at or very close to therapeutic grade. Which means like they have to bind really tight.
37:15Speaker C:They also have to be developable. They have to have like all of these nice therapeutic properties and developability.
37:19Speaker A:I think the, we talked about, he mentioned Chai 2.5.
37:23Speaker A:Which we released like a few months after Chai 2. There was a study we did on the developability of the molecule which you know, for the audience, like obviously the molecule has to stick good and stick tightly but you know, there are these other properties you care about and to use the non biological terms. Right. Is it, is it safe, is it stable, is it easy to manufacture?
37:41Speaker A:Does it, you know, self aggregate? And we've, we've been pleasantly surprised at, you know, how, how much we've been able to climb and push the performance in those areas.
37:50Speaker D:It seems like one of the reasons that you want to do antibody is because the developability.
37:56Speaker A:Yeah, you get a lot for free there, right. With that antibody framework.
37:59Speaker D:Yeah, it's interesting. I mean to me there are many structure prediction molecules out there. I mean models out there. I feel like the, it's these other ancillary factors actually that are going to probably be the most impactful in the usefulness of a, of a product.
38:17Speaker C:Yeah, right, yeah, absolutely. The nice thing about structure prediction is there is a ground truth that you can compare against for design. You don't really have that. You're like, here's some new like disease molecule, give me a binder for that.
38:30Speaker C:And like, if you want to know if this thing really binds, you have to send it off to the lab and wait a while. For structure prediction you can be like, all right, the model hasn't seen this sequence before. It's never seen anything close. Does it actually fold up into the correct shape?
38:42Speaker C:And we can just kind of hold that out of the data set and check. So I think I've always thought of structure prediction as this really nice speed run kind of benchmark to like validate ideas on.
38:51Speaker D:Right, sorry, I didn't mean to say, I meant, you know, sort of structural models in general. Yeah, but yes, exactly. So maybe we can talk a little bit more about, start getting into the product side of things. Thank you for coming.
39:06Speaker D:I actually, I mean like, like I said, I really think this goes throughout not only for, you know, sort of structural models like this, but also VirtualCell and whatever. It's really the, all the other stuff around the, the direct development process that is going to have the biggest impact. So you talk a little bit about that.
39:25Speaker A:Yeah, I think that's actually a good thing to talk about after Chai 2 because I think Chai 2 is where it started to get really fun from a product pers. Right. I think with Chai2we, we crossed the threshold of usefulness where after we, you know, released that paper, we had a lot of, you know, you know, pharmas and biotechs approach us and say, hey, this model might be able to do some stuff for us, like can we use it? And then we're like, oh man, like we, we should build a product, right?
39:48Speaker A:We should build something to let you use that model. And that's right around when I joined and there was sort of this, you know, mad, mad build out to both, you know, build the product which we can talk about the shape of and also go and secure the compute actually so we can go and serve those models to our partners. And you know, I think another third piece there that was really interesting is, you know, around security and ip.
40:09Speaker A:I think we want to be a very neutral platform that anyone can design medicines on. But as you guys know, like pharma is this notoriously IP sensitive industry. Right. And I think when I joined a lot of people told me this can't be done.
40:21Speaker A:Like they're not going to put their data in a platform and like have all their new medicines be generating out of it. And having a bit of a background in security helped a bit. Whereas, like, no, actually, if you, like just are really aggressive about how you like segment data and set up like single tenancy, where you're like almost deploying a separate version or a separate account in the product per customer, you can actually build a platform and then go and ship it to them. And so through the summer of last year, we started doing that, right?
40:49Speaker A:And we'd been working with or talking to Eli Lilly and they were one of the first partners to really work with us closely on that, kind of made that V1 of that design, design suite that you can use to engineer some of those molecules on. And maybe it's worth talking a bit about that design suite. I think we have these really, really powerful models now that can do all of these crazy things if you condition them in the right way, if you kind of give them the right context about the structure that you're going after, or maybe the constraints around the model, like, hey, I want to design an antibody that hits this GPCR protein but, you know, doesn't collide with the cell membrane and also targets the specific epitope on that as well. And you know, we looked at that, we're like, I guess we could put a chatbot around it that'd be like really easy to talk to, but like, really, like you're trying to build something almost very visual, right?
41:44Speaker A:And you can finally build something really visual with some of these structure prediction models. And so if you kind of look at the CHAI product, it looks a lot less like a ChatGPT and a lot more like Autodesk or SolidWorks or Figma. You know, if you've used those things where you can kind of load up your molecul, there's this almost like Photoshop esque design suite. You have this equivalent of a paint tool to kind of paint your epitope.
42:08Speaker A:You have this equivalent of a content aware fill tool to kind of get your binders generated from chai. You of course, have a lot of the scientific analysis and plotting and whatever to understand the results of the models. But we've just been surprised at how much complexity is actually just in doing that. Right.
42:25Speaker A:So that you kind of don't shoot yourself in the foot when you're then prompting these models to give you binders.
42:30Speaker D:So are you sitting with people who are designing These antibodies, you know, and like, and then they're complaining to you or whatever.
42:39Speaker B:Yeah, how does that work? How do you convince med chemists to use your tools? Because med chemists hate AI tools. Like, notorious.
42:47Speaker B:Like, I don't want to touch this thing. Or, like, I don't understand it. And they will not touch things which they do not understand.
42:53Speaker A:Well, it helps a lot to have the models working really well.
42:55Speaker A:So when we, when, you know, when we had the results of Chai 2 and Chai 2.5, I think, you know, that's enough of an activation energy where, you know, pharma companies and scientists within these companies are like, oh, let's try it. Actually, can Chai. Can you guys just try running the model against a few of these targets and let's look at the results. And then we do that, and the results are good.
43:15Speaker A:And they're like, okay, let me, let me try to get on that product and let me try to use it.
43:19Speaker C:No, I think pharma is, like, incredibly pragmatic, actually. Like, I've been very impressed with everyone that we've, we've worked with so far. They're, they're very, like I was saying, pragmatic about this. And they're like, they're, they're willing to be proven wrong.
43:31Speaker C:And like, I actually don't blame them for not trusting the models. Like, I've used these models and like, so many times. They've, like, rightly so, like, I, I, I am pretty skeptical when I, like, see any release. I, I always have been.
43:42Speaker C:So, like, you really just, like, need to show them the proof and, like, they can give you this target that they are interested in. Or maybe it's more of a sign they've worked on in the past. They probably don't want to, like, share IP right out of the gate, but they can be like, hey, you know, I've had trouble with this particular target in the past. Let's see how, how you guys can do on this.
43:58Speaker C:And then once you show them the proof, they, like, almost overwhelmingly are willing.
44:02Speaker A:To accept that I come from a cybersecurity background or, you know, have worked on security products before, and those were dark, dark years. Because you spend a lot of your time actually selling to people who are surprisingly not that technical. You think cybersecurity people are very technical, and in many cases they are not. And it is this kind of, like, uphill enterprise slog to this very unsophisticated customer.
44:22Speaker A:I think we've been just pleasantly surprised, or I have by Just how much I enjoy working with our partners and our customers. You know, these are scientists who have been spending, you know, 5, 10, 20 years of their life working on one target. Right. Often in some cases.
44:37Speaker A:And they've studied everything about it. You know, they're, they're very sophisticated, they're very smart. Right. You know, getting to collaborate with them is just a goldmine.
44:45Speaker A:And we learn a lot about how to make the product better. You know, this anecdote, we, you know, a few months ago, we were actually showing some of the results that we. From a Target that with a pharma partnership. And one of the scientists in the room, like, started tearing up and crying.
45:05Speaker A:And she was like. We were like, what's wrong? She's like, no, I've just been. I've literally spent 10 years trying to get an initial binder to this thing, and you guys were able to help me do it.
45:15Speaker D:Oh, that's.
45:15Speaker A:And you know, that. That feels really special. To answer your question, you know, we, you know, there's of course, the teams of scientists and computational biologists that we're working with within each of our partnerships. There's also the people we have within the building.
45:28Speaker A:Right. So I think one of the things that I really appreciate about CHAI is how cross disciplinary it is. We have people who are maybe engineering experts and less bio experts like myself. We have great AI scientists or ML scientists, but we also have a bunch of scientists that we work with and have joined CHAI to sort of help us both test the limits of the models.
45:50Speaker A:Right. See, what is Chai 2 actually capable of? What targets can it do? What can't it inform some of the research direction there?
45:57Speaker C:I want to add to that, like in, in like the CHAI two days, like, we, we kind of started with like a bunch of engineers and people who have like AI bio experience. We didn't have a hardcore lab scientist. And like, one of our first hires on that realm was Nathan Rollins, who I think he, he started working in the Baker lab at 14, graduated from Harvard at like 18, and got his PhD by like 21 or something like this in the Marx Lab. And he was like, super skeptical about CHAI at first.
46:25Speaker C:And then, you know, the results start to come in. He's like, okay, this is, this is kind of interesting. Like, this could work. And then like, once the Chai 2 results came back, he was like, I need to bulletproof this.
46:34Speaker C:Like, nobody celebrate yet, like all this. So I think, like, it's. It's been really nice to have that level of rigor and to just have people who have really, like, they've spent the time in the lab, they've designed proteins themselves, they've literally, in the case of like, Andy, led several therapeutic programs, brought drugs to the clinic themselves. And like, we have all these people internally at CHAI just like using the product and like really battle testing that.
46:57Speaker B:So if you don't have your own platforms, right, I mean, so you don't have your own programs, right? You're a pure platform or partnership model, right?
47:04Speaker C:Yeah, yeah.
47:05Speaker B:How do you battle test something if you basically aren't. You don't have a use case where you have to continuously push it forward, or if you are just pushing things forward, when you just end up with your own candidates, if you're successful. And then what do you do about that?
47:18Speaker A:I mean, we have benchmarks of our own internal cases, right. You know, there's a set of targets that, you know, have. Are known therapeutics, right, that have known therapeutics against them. There's a set of targets that we pick to sort of push ourselves, right?
47:29Speaker A:And so we're constantly refining that set and adding to it. And that's what that internal science team that we have helps with, right, Is expanding that and almost running the experiments to try to get initial binders there. We don't care about going and developing those drugs. Like, we just do that in service of validating and making our models better.
47:46Speaker A:And then of course, there's a loop with, with our partners too.
47:48Speaker B:Would you consider yourself hit discovery or are you. Do you, I guess using some jargon, hit to lead. Lead optimization, like, where do you live in this? And you know, hit discovery might be like one part of it which you can do hit discovery.
48:01Speaker B:But the, the later, the, the other parts of this are, I think, oftentimes much more bespoke and kind of special. I mean, how do you balance that? And it's. It seems much more, much more difficult to me to be general than it does to.
48:15Speaker B:To solve general lead optimization than it does to solve like, hit discovery.
48:19Speaker C:I think ideally we really want to be able to. Rather than think of this as a bunch of stages. I think part of the reason why we think of it that way is because the initial molecules are usually not good enough to be drugs. And really we're kind of at the inflection point now.
48:35Speaker C:We're really seeing this internally at chai, where the models are getting pretty close to producing molecules that could eventually or like, are very close to drugs. So we, we try not to make Too much of a distinction between, okay, hit discovery, lead optimization. All of the different parts of this kind of preclinical pipeline are, are like, you know, the, the light, the North Star is to just really produce drug like molecules straight out of the models. Of course this is going to be hard and like they're going to be like tons of roadblocks and like you need to be able to like actually prompt the model to do this.
49:06Speaker C:You need the whole RL stack to like learn different properties, things along those lines. But I think it's very achievable. Yeah.
49:13Speaker A:And I think to add to that, right? Yeah. This notion of target discovery and hit discovery and optimization, where each of these has a gate and takes a few months to a few years, is this very waterfall model, right. Where the cost of trying things and getting things early is very expensive.
49:27Speaker A:But I think to what Matt's saying, right, if you start to get in a regime where you can have models give you really promising candidates, you can start to make that look a lot more like a loop, right. It's akin to like becoming more agile and software development. Internally we kind of have two North Stars, right? And at first pass they almost sound like contradictory.
49:46Speaker A:But the North Star in research is to start to de novo one shot better and better and better medicinal candidates that are as close to being ready for the next phase as possible. But also within product we do want to expand into whatever these iterative workflows look like, where maybe I get a binder, I get some results from the lab, I'm using that to condition my next run of the model. And I think, you know, they sound contradictory, but I think they're actually not. Because I think what's going to happen, you know, the research is going to get better at identifying a de novo candidate for like a specific class of drugs, right?
50:21Speaker A:Say like antagonists, right? Like blocking things, right. Little bit easier maybe. Okay.
50:26Speaker A:We can get to a state where we can one shot, pretty good drugs there. But now the next problem is like agonists, right? Like how do you reliably. One shot hitting a switch like on a cell, right.
50:35Speaker A:Or bispecifics or ADCs. Right. And I think, you know, there's kind of this levels of abstraction that we're going to have to climb with the product as like the models get better. One of the things I was, I got, I got very existential like a few months ago because I was like, man, all this stuff we're building in the product to like visualize molecules and do this, like, maybe I'm just going to have to throw it all away when, like, Matt ships like Chai 4.
50:58Speaker A:Right. But, you know, I think that's. That's kind of the reality of like building products now, right? You're actually using them less as an end in and of itself.
51:05Speaker A:Like maybe you'd have built software that was supposed to last like 20 years. Now it's supposed to last maybe one. But it is the bridge to deliver value and kind of enable the research that then gets you to the next thing. And so I'd imagine we're probably going to rewrite our products at higher and higher levels of abstraction, right?
51:21Speaker A:Like maybe like right now we have something a little bit more akin to cursor, where you're, you know, inspecting the molecule in the same way you're inspecting the code because you really need to verify, like, the bonds that are forming and the properties of the things that you're getting. But then, you know, you get to a point where that stuff is solved enough, where now the product is actually just helping you orchestrate these campaigns of hypotheses, right? Or maybe you have one target and you're orchestrating a bunch of different epitope choices or whatever against that. And then maybe you're going up one level of obstruction where you're now doing a whole campaign against all of the targets within a pathway, right?
51:55Speaker A:And I think what's really exciting about that is if you have these really good primitives for structure prediction and binding and design and you can kind of compose them, then you can start to just like, grow into, like the outer loop of science, right? And then, you know, maybe the thing runs itself and you start to really get to some really, really, really cool drugs at the end of it.
52:14Speaker B:I actually want to push on what you just said about epitope prediction because I think a lot of people in the field would argue this might be the much harder problem than finding antibodies and binders. Where do you think that the state of the art is in general and also with regards to Chai in terms of epidote prediction and like, is this a problem which has a reasonable solvable time hor. Oh, and also maybe can you define epitope prediction?
52:37Speaker C:I'll think of this at, like, some different levels. So the most basic level is, okay, I have some disease that I want to target and what proteins are actually responsible there, like, actually figuring out biologically what's going on, like, what should I be targeting in the first place with the drug? Because once you figure that out, it's kind of like a structural biology problem. At that point, you're like, all right, this, like, set of proteins is responsible and, like, what's going on there?
52:57Speaker C:Well, this is interacting with some other protein that it shouldn't be interacting with with. And conventionally you'd just, like, want to block that interaction or something with anybody. But that's kind of where these proteins interact and like, the type of interactions that you want to disrupt, that's typically like the epitope. It's like the actual site on the protein that you want to block.
53:14Speaker C:This is a ridiculously hard problem. I'm with you on this. This is like the harder problem, like, just the amount of context that you need and like the global understanding that you need, you need to get in order to, like, actually figure out what's interacting and how.
53:28Speaker B:But maybe let's take a few specific cases. Let's think about, what about SARS CoV3 comes around or the new flu or whatever. What would you do there? I mean, is that something that you think you could actually reasonably tackle in that case?
53:41Speaker C:Like, yeah, you could just run a structured prediction model maybe and, like, see where the model thinks this thing will bind. If it's highly confident in that, you might say, okay, here is like, the site that we want to block. I think, in general, still very hard. And even, like, structure prediction, it's, it's getting really good.
53:56Speaker C:And like, a lot of people think Alpha 2, like, solves structure prediction. Not really. Like, Alpha 2 got, like, I think, 11%, the multimary version of this got like 11% of antibody antigen prediction cases. Correct.
54:09Speaker C:That means 90% of the time it's wrong.
54:10Speaker B:Yeah, yeah. I mean, but, but AlphaFold2 solved a certain class of monomeric proteins with MSA.
54:17Speaker C:Absolutely.
54:17Speaker B:Yeah. Yeah, yeah. So, I mean, the, and that's the msa, I think, might be the, the key point here because MSAs are sort of the, the, the magic which makes it all work. It's like a, it's a template in some sense, about, like, what the structure should be.
54:30Speaker B:And antibodies almost evolutionarily can't have a template. Right. Yeah. Everyone has to have unique antibodies custom to the things that they've experienced over the course of their life.
54:40Speaker B:So. Right.
54:40Speaker D:And just, just to clarify, I, I had to understand this myself, so maybe I can help the listeners who aren't familiar. An antibody, the whole point of an antibody is it. It can identify. It could be used by the immune system to identify new things that it hasn't.
54:53Speaker D:The body hasn't encountered before. So the design of antibodies as opposed to other types of proteins is to. The system is designed so that you can quickly recombine different components of it in order to match proteins that are from unknown pathogens, more or less. And so this is why you.
55:13Speaker D:It's not conserved in evolution the way that other parts, other proteins are.
55:18Speaker C:Yeah, so, so like back to the, the apto prediction problem. I think it's still hard. I think like there are a lot of cases that are maybe tractable, but I think in general, like if you want to discover this for a new target, still a really difficult problem. Maybe VirtualCell would be like the closest thing to state of the art there, but that's still a ways out.
55:37Speaker D:I wanted to dig in a little bit on the product because there's something I don't understand about the economics of basically all the structural stuff that's happening right now. And obviously a lot of people think it's very, very valuable. So there's, you know, I'm not grokk. When you look at the cost of developing an antibody, you know, it maybe is a couple million dollars, right?
56:01Speaker D:When you go from, you, you, you've identified a target somehow and then you say, okay, I need an antibody to match this and then I have to sort of optimize it in various ways and then maybe I try it in. I mean with antibodies you go to animal typically faster. If you look at how much does it cost to drink, bring if you like are prescient and pick the right target and the right great technology to get all the way to drug, it might be half a billion. Typically that $2.6 billion number is amortized over all the failures as well.
56:32Speaker D:So if you look at just the cost of that one success, depending on the, the disease, maybe less, but you know, half a billion might be a good median number or something. So you're, you're saving like a couple million dollars in a half billion dollar campaign. So why is this so attractive?
56:50Speaker A:I would maybe challenge the premise a bit like in a few ways, right? Like, okay, sure. If you're trying to get an antibody for like a very simple kind of target, like maybe. Right.
56:59Speaker A:But I think what we've been most excited by is our partners using antibodies in, you know, more sophisticated ways. Right. Like in, for example, in Chai 2 we showed like GPCR agonist activity right. Where you can really hit the switch on a, you know, on a cell doorbell protein, so to speak.
57:14Speaker A:Right. In a very precise way. Very, very, very hard to do. That with antibodies, if you can't be that precise.
57:21Speaker D:So you're unlocking a new capability.
57:23Speaker A:I think about it as less like, oh, I'm taking the ex that I can do and making them faster. I mean there is some of that too.
57:29Speaker A:But it's like, no, they're just like, hey, how do you go after like better targets?
57:33Speaker A:That are, you know, maybe more precise, more effective. Right.
57:37Speaker C:I think like also on top of that too is like there are drug modalities that you just can't discover with immunization. Like you're not going to design your like crazy multi specific warheaded, super intense formats. These are really things where you kind of have to design these from first principles even just, just with bi specifics in particular, like both arms need to now bind different targets and you've kind of like have this multiplicative effect on your binding rate. So like if you have a one in a billion chance of finding a binder in arm one and a one, a billion chance in arm two.
58:08Speaker C:Yeah, you're not. This isn't going to work with the traditional approach. Exactly.
58:13Speaker A:I think the other thing I'd think about is, right, you're not just helping your partner with maybe one drug. Right. There might be a portfolio of targets that are, they're going after a portfolio of drugs that they're trying to make. And, and the nice thing about the platform approach rather than the we are developing individual drugs is we can sort of scale with them as they pursue more targets in addition to more ambitious targets.
58:33Speaker D:So it like lets you concentrate your, your learning in a. Yeah. Sub domain of that and so that you. Everybody benefits from that. Exactly.
58:42Speaker D:That's the. But I, okay, so I didn't. So what is, what are some of these capabilities you mentioned? A few.
58:46Speaker D:Are there more that are really interesting that you guys are chasing? Yes.
58:50Speaker A:I mean we talked about like, you know, cross reactivity, we talked about selectivity, we talked about some of these like really interesting additional modalities with bispecifics. Right. There's a set of things that, you know, our partners have been asking us for that we've been working on that I can't get too into because then that starts to reveal some of the, the targets that they're going after. But I think that the point being you can just, once you get precise, like you can start to do some really, really cool drugs.
59:16Speaker D:It's a new technology.
59:18Speaker D:So like the technology in, in pharma means like how do you deliver your therapeutic and so this is maybe a kind of thinking about like car T is a technology. Right. And so this is maybe a new technology in the sense that you can have these highly, highly engineered.
59:36Speaker A:Right. And that comes, you know, from the mission of the company is to really turn, you know, drug discovery from a scientific experiment to an engineering discipline. Right. How do you sort of get to the precision engineering ph for biology where you can start with, you know, almost declaratively define the thing you're trying to get and have the model fill in the gaps and get you that.
59:56Speaker B:So what is the biggest blocker from going from science to engineering?
1:00:02Speaker B:That's your goal.
1:00:02Speaker A:There's so many things like that's the thing about, you know, micro heterogenicity.
1:00:06Speaker D:Yeah, micro heterogeneous.
1:00:11Speaker C:I don't even want to talk about this. Like the amount of headaches, like too late.
1:00:14Speaker B:You already went. You're kidding.
1:00:16Speaker C:Okay, so like, just, just like when you're actually parsing like first of all, file formats for biologists, like, I just, they just don't care. There's like no standardized. There, there are standardized file formats. Are they the best?
1:00:27Speaker C:I. I don't really know, but there's like also just like a lot of information that you want to pack. And I have the structure. Here are the people who solved it. This is the method I use to solve it.
1:00:34Speaker C:There's like a lot of stuff going on. And then depending on the method that you use to actually figure out what this 3D structure is, you might have like multiple copies of that structure. Part of it might not have really been resolved. Or you're like, it could be here, it could be there.
1:00:45Speaker C:I'm just going to give you both options. So the actual just parsing problem on the engineering side of working with this type of data is really difficult.
1:00:53Speaker B:This seems like something that LLMs can.
1:00:55Speaker A:Excel at, though they don't know all the edge cases often.
1:00:58Speaker C:Right. This is more back to just a simplicity approach. LLMs are very good, I'll absolutely give you that. Then you're thinking about, do I really want to.
1:01:06Speaker C:Should this function have 20 special cases or should be really principled and how we approach this? And should we be, I guess, more of a. Opinionated. Opinionated, yes. Like how opinionated should we be and how we do this?
1:01:19Speaker C:We want a strategy that's easy enough for humans to understand. And like when we're reading through the code base, we really need to know what's going on here, what are the potential problems. And like, sometimes that just comes down to looking at examples. But then I think, okay, once you've kind of figured out all the info work and how you get data into the models, there's then like scaling the model model, there's then scaling the infrastructure around the model to train bigger and bigger versions of this.
1:01:42Speaker C:And that's like a lot of work that Neil and the product team actually.
1:01:45Speaker A:Yeah, I mean that would have been my answer is the infrastructure part. I mean, you know, not to beat a dead horse, but compute, right? Getting the compute and using it in the right way is such a challenge. You know, it's especially for startups and.
1:01:57Speaker B:This has been such a theme. Yeah. Anthropic is single holding back science.
1:02:05Speaker A:To that point.
1:02:07Speaker B:I mean they're also accelerating science. But it's like this weird. Totally.
1:02:10Speaker A:Like one of the, one of the things that I help a lot with at Chai is buying compute for the company. Worst job, man. I would not recommend it. It is very stressful.
1:02:20Speaker A:But you know, even September of last year, right back to you.
1:02:23Speaker D:The hardware job.
1:02:24Speaker A:Yeah, yeah, I know, exactly in the wrong, wrong way. But you know, September of last year we started to really notice like things are getting, getting tight, right. We were doing a lot of our inference on, you know, spot and on demand markets and we'd have these days where you just like get these capacity crunches and we're like, okay, we should probably start to get ahead of buying some compute for ourself. And I mean, I think everyone probably says this, but man, it was, it was hard.
1:02:50Speaker A:Like I think I didn't, I didn't realize how much of a power law, you know, this is, right? Where you know, there's, there's say 10,000, you know, B300 units that are shipping everywhere, right? The, the hyperscalers and the, you know, the biggest, the big AI labs are buying 95 plus percent of it, right? And then you kind of have the startups like fighting over the scraps.
1:03:11Speaker A:And I think the other thing that's really interesting, especially if you look at these later compute versions, right? The Vera Rubins or you know, the B300s, like a lot of this stuff has been built very like LLM for it, right? Like you have these systems with like huge KV caches where you have like 72 GPUs that are all acquired to talk to each other, right? And obviously some performance gains there help us.
1:03:33Speaker A:But it's kind of interesting just how much the compute market has kind of gotten LLM pilled. I think there's a whole probably set of Compute, stack and inference optimizations and things that need to be made for this class of models. And I think this class of models is going to be just as big, just as impactful as LLMs. But it's almost like the compute market kind of doesn't realize that yet, both in the capacity sense, but also in the software stack sense.
1:03:59Speaker A:So we actually spend a lot of our time even just doing basic optimizations of compute to get them to work better for the types of models that we have.
1:04:07Speaker D:Yeah, I know that some structured models are more recursive than LLMs, for example. And so that. Which changes sort of like maybe the compute to memory ratio that you need and things like that. What are some of the sort of cool or interesting optimizations that you've done.
1:04:26Speaker C:There, depending on the type of model. So like, we can go back to like a chi1 type model. In that case, we're following the alfold 2, 3 architecture. And there you're like, rather than doing attention over like this like normal sequence representation, you're in a sense loosely doing attention over this pair representation.
1:04:44Speaker C:So you can think of this as like a sequence of length L squared rather than like typically length L. If you're doing attention over that, the way that you actually batch this up, it ends up being L cubed. Now you're in a pretty heavy compute regime. So the amount of flops that you're putting into every token stays. It's pretty high.
1:05:01Speaker C:The amount of memory that, the memory bandwidth overhead of just transferring that from SRAM to whatever, that's a real bottleneck in these architectures. So even something as simple as a layer norm can take a long time actually, that can be a significant amount of the compute that you're using. So I think on our side, we've spent a lot of time just optimizing and engineering, taking engineering very seriously so that operations are, you know, at least better. We're always looking at like, how do new chips perform compared to the older versions.
1:05:32Speaker C:Sometimes that's even different for training versus inference. And like, of course, Neil knows this really well.
1:05:37Speaker A:Well, so, so there's, you know, what you're doing on the individual gpu and then there's like, how do you like orchestrate fleets of GPUs, right? And you know, you basically shard your computation, right? And so, you know, when you're designing a molecule on Chai, it's not necessarily like one call, right? It's a lot of, a lot of GPUs being thrown at the problem, right across, across a lot of compute.
1:05:57Speaker A:And actually I would say that one of the hardest things to get right in software engineering is durable execution. Are you all familiar with that term? Can I go on a little? Ultimately, if you're computing a lot of data model calls across a very wide set of infrastructure, you always run into these problems where some part of the infrastructure is flaky.
1:06:20Speaker A:Maybe the bucket you're grabbing your data from goes down or your database has a blip because there are too many transactions against it or you have GPU errors out, right? I've been at companies before where you like spend so much of your time just dealing with this shit, right? Like you, you're basically putting like all of these queues and like all of these retries and you're like duct taping things together and you have a, and it becomes this mess where now what used to be like a, ideally a pretty simple like computation that's just distributed, you're ending up spending like 95 plus percent of your time on all of this queuing and retry stuff, right? We're huge fans of this company called Temporal.
1:06:57Speaker A:Basically there's this idea like, look, if you're just trying to get something, a really long running job to run at the end of the day, what do you need? You need a queue, you need your flaky thing pulling off of the queue. You need some retry logic to put things back on the queue if they fail. And then you need some whole orchestration system to just tie all the queues together and monitor them.
1:07:18Speaker A:What's really cool about Temporal is this is a company that's kind of invented a framework for doing this. And, and I think one of the technical decisions we made early on that was very helpful was to run as much stuff as we can on Temporal, right? So whether those are calls out to the database from the app, right, to make sure the database transaction goes through without failing. Okay, let's have side effects sit on Temporal so that they get retried smartly without us having to write our own queue logic, right?
1:07:45Speaker A:Or things related to model calls or things related to orchestrating really long data pipelines. Point being like, you know, one of those primitives like just like, hey, you need to get durable execution, right, so that you're not stuck in like retry hell. A really deep like engineering thing that like you wouldn't realize if, unless you for like me and Jack, you've been like burned by this like many, many times before. And I think like we're at this state now, right, where we've you know, we've raised another $400 million.
1:08:13Speaker A:I have to go buy another compute cluster. Like, you know, like we're going to have like really, really, really large runs and inference and training sets. And so getting those foundations right is what's actually going to let us do more ambitious things. And to kind of answer your question, I actually think that's a lot of the bottleneck to making biology more like engineering is just like having the right engineering primitives supporting it.
1:08:36Speaker C:I have an analogous tangent on the model side. Actually one of the things that's kind of nice about those problems is they're super visible. So at least, you know, hey, this crashed, this failed for us. We just see like loss curve didn't go down or we see weird gradient behavior or whatever.
1:08:52Speaker C:I think think a lot of these same principles like you know, engineering first. That also applies on the research team. One thing that I like to say is kind of like complexity and being bitter lesson pill. They're like fundamentally at odds.
1:09:02Speaker C:For example, I think like alfold3 I might get this number wrong, but I think it was like 23 sub modules. And at that point that's a really difficult system to optimize and study. You're like, all right, what happens if I change like if I tweak this thing in sub module 30 or like 21, what, what happens to the whole system? And you can always think, think hey we can make this better by like adding module 24.
1:09:23Speaker C:But like should you, or should you think about just like removing things and lowering that complexity down? But I think that's like a pretty fundamental thing at CHAI is just like the engineering culture and just being like very simplicity biased.
1:09:33Speaker A:Have y' all seen the picture of like the SpaceX engines? It's like Raptor 1, it has a bunch of pipes and like Raptor 2 we have a picture of that like on our office wall. Because I mean it's just true, right? Like how do you delete, delete, delete more things?
1:09:46Speaker B:Yeah, but the only way you can accomplish that is, I mean the reason AlphaFold2 and AlphaFold3 worked, they were small models, relatively speaking. They were very compute intensive but they were very data efficient.
1:09:59Speaker B:And like the, there was inductive bias after inductive bias.
1:10:03Speaker B:Brought in by human intuition and probably like hard fought experience. It was, they're incredibly efficient. If you try to, to knock down those things, you know, they're not like a house of cards. Like everything is a incremental improvement on top of it.
1:10:21Speaker B:In order to get beyond that, it seems to me like you really just need new sources of data. You would need to at least treat data fundamentally different in a way that is much more efficient. I mean, I mean, I'm actually kind of surprised to hear that you have scale to that degree because it suggests that you're doing something very different from what the community is thinking, the way the community is thinking about it. I don't know if you can comment.
1:10:43Speaker C:About that, but we're pretty first principal people, like the whole research team at CHAI, except for me and Kevin, really, like we're the only people with quote, bio background. Even still, we're pretty far removed. So I think we try to look at every problem as a core ML problem. We try to think of what's the analog in other spaces.
1:11:02Speaker C:So even for image models, CNNs were built to process images. Images should be looked at in patches. That was the nice inductive bias that there. Then people are like, well, you can just kind of tokenize this thing, throw it into transform and it's going to work.
1:11:15Speaker C:And like it did end up working even like on a relatively small data set. But I think for proteins in particular, it is really hard. There's not as much structural data, there's a ton of sequence data. Like that's one of the unlocks for like ESM working.
1:11:28Speaker C:You can get that to just run on a transformer. If you try to do the same thing with like experimental structure data. Good luck. You need alphafold.
1:11:35Speaker B:Yeah, I mean there was the, there was that Apple paper where they distilled on the alphafold, which it was actually really cool that you could distill on a very large data set and you could get, you know, good signal. But, you know, it didn't generalize at all because it wasn't reasoning, it was really pattern matching. Yeah, like one of the things, these like triangle layers you were talking about, for example, they do have a very nice inductive bias. Maybe it's not the triangle inequality like the paper originally proposed, but it's a clean inductive bias and it unambiguously is like one of the things which made it work.
1:12:06Speaker B:And it just comes at a huge cost.
1:12:08Speaker C:Yeah, yeah, no, I think that's, that's definitely true. These layers are pretty costly and like that kind of limits what you can do with the architectures. They're not like, not only are they like costly in terms of compute, they're just like not efficient on modern GPUs either. You have small hidden dimensions, large sequence dimensions.
1:12:24Speaker C:Like it's like exactly the opposite of what GPUs are designed to process. One takeaway from like triangle layers is you're kind of just trading off parameters for compute in that sense. Like, that's like one mental model for thinking about things this. I might want to throw more compute at the problem and just trade that off for parameters because I won't be able to hold as many.
1:12:42Speaker C:I can't literally store these large pair representations and still do normal attention. So I think there are fundamental things you can abstract from the ideas, like AlphaFold, but you can kind of just tweak these and start building off of them in your own way.
1:12:57Speaker B:It sounds like you have quite a bit of research, like fundamental research going into this direction for, I guess, audience looking for a nerd snipe in ML engineering for new problems. Probably something very. It's a very different research direction than a lot of the communities going in.
1:13:14Speaker C:Yeah, yeah. I think what we built at CHAI is like, it's, it's very unique in a lot of ways, but also very tied to like what core ML is, is good at. Kind of what I was saying before, like, we try to map every problem into like a core ML problem. We think, you know, how would you approach this if, if it were an LLM or something like that?
1:13:31Speaker C:But yeah, like, at the end of the day, we really, really value simplicity and we, we really encourage people who don't have a bio background to like not be scared of this stuff.
1:13:41Speaker A:And I think that extends into the product too where, you know, there's a balance to be had here, right, between like, how general do you make the product? Like, do you build a cross reactivity workflow and a selectivity workflow and a bispecifics workflow? Or do you all say, no, like, let's make the model general enough to say, I'm going to like condition on arbitrarily binding or avoiding something. And then you just have a very, very general screen in your CAD suite where you can say, hey, I just want to avoid or bind to these parts of these different structures.
1:14:08Speaker A:Right? And I think, you know, kind of like the ML team, like, I don't, I don't have, you know, a formal BIO background. Most of the product and platform team doesn't have a formal background either. Now there's some amount of like, maybe regretting my words that I'm going to have, right?
1:14:22Speaker A:Because I'm sure there are, you know, a million nuances and, you know, don't want to come off as, you know, Too brash or naive there. But, you know, I think sometimes it's helpful to not be burdened by, like, all of that, the. Oh, these, this nuance and this nuance and this nonce.
1:14:34Speaker B:And you can.
1:14:34Speaker A:You get to kind of bet and be maximally general because, you know, that's kind of what we're seeing in the research. You can. The models are very general. That lets the product be very general.
1:14:43Speaker C:I'm thinking back to, like, in my. In my CS theory days, my first advisor was like, we're working on some problem and we. We need like, a polynomial time algorithm for something. And he.
1:14:53Speaker C:He would always tell me, like, never underestimate the power of polynomial time. Like, this is basically like, you're allowed to choose, like, whatever exponent you want. And my first paper was an end to the 20th time algorithm for this problem. And I was like, andy, I did exactly what you said.
1:15:06Speaker D:He's like, wait a minute, I didn't mean it like that.
1:15:08Speaker C:Yeah, I think, like, it kind of, like, you can really help yourself. Like, you can free yourself up a lot when you're like, all right, I can kind of do whatever I want and then kind of simplify it later. And I think that's really, like, a pretty fundamental way of thinking about things that we. We leverage a lot at ch.
1:15:23Speaker B:The space of binders, of protein design and binders in general is actually a fairly crowded space. I'm curious about what your general outlook of the field, the industry is. I mean, I can go back to, like, some anecdote. I was at maybe NeurIPS 3, 4 years ago, the one right after RF Diffusion came out.
1:15:42Speaker B:I was talking to someone in the Baker lab, and they're like, man, I just one shotted. I don't think they even use one shot. One shot wasn't even a term back then, but they was like, I just got pico molar binder out of RF diffusion and just like doing the cryo. Great, right?
1:15:57Speaker B:It didn't seem like that just solved the problem. Like, it's not like, oh, man, now every. Yeah, but there are lots of people who I think have seen that you can actually do protein design, at least in some categories, quite well. I'd say, like, is it mini proteins or mini binders?
1:16:15Speaker B:Ironically, nanobinders are actually smaller than or larger than mini proteins. Genes are maybe, like a little bit harder. Antibodies are typically considered even harder. But there's this, like, is this something which can and will be commoditized, at least in some part?
1:16:29Speaker B:How do you compete like where does this, where do you, where does the field go from here?
1:16:34Speaker A:I mean, I think the answer is it's kind of all of the above. Like I think there probably will be some commodity layer for, for certain types of modalities or drugs, right? I think at the same time we're going to be able to do even more and more and more ambitious drugs and you're going to. It's just like what's happened in LLM land, right?
1:16:51Speaker A:Like you have your, your open source models that are maybe general and helpful for some things, but people are still buying frontier models, right? And actually if you look at the amount of value captured, it's actually the closed source frontier models. You know, the whole pie is growing, but it's growing so fast that even as the open source models like SHARE expands that the, the frontier models are still able to capture the majority of the value, right?
1:17:13Speaker B:Raise your hand if you're using an open source model on your day to day, right?
1:17:16Speaker A:And what are for that? Right? One like if you have, you know, more intelligence, you're going to go after harder tasks, right? I think if we have more, you know, intelligent bio models, we're going to go after more, more crazy bio tasks, right?
1:17:30Speaker A:But then also too like, I mean a lot of the reason I don't use the open source model is because like you know, I don't get like cloud code, right? I don't get like cloud. You know, I think there's a product layer to be built that is just as important as the model layer. We learn a lot from our partners and you know, the people in the building as well.
1:17:46Speaker A:Just like what are the really, really tough things that they get stuck on using the models, right? And some of them are like, you know, the dumbest things, right? Like you know, I want to be able to better visualize this piece and like focus on that. And some of them are actually like very sophisticated things that we then have to build some like pretty vertical product for.
1:18:02Speaker A:And look, maybe in the fullness of time like AGI, like one shots, everything and doesn't matter, but I think there's quite a bit of time until we, we get there, right? And I think the, the product makes a huge, huge difference for that. That'd be my answer. I mean you probably have a more model forward answer.
1:18:18Speaker C:No, like I think like biology is slow which is like one kind of nice thing and there's like not that much labeled data. So like you could take all the publicly available sequence information out there that might give you a Good base model but you still need some measurements on that data. That's still pretty time consuming and then you need to like iterate on that. So I think there are even just data blockers there into unlocking like if we really want to do this zero shot design candidate, start generating molecules that are almost ready to go into the clinic.
1:18:47Speaker C:I think it's more to that than just like, you know, AGI might not solve that right away. I think there are definitely like some technical blockers there.
1:18:55Speaker B:But even in the space of you know, specialist companies, I mean I'm not going to like to start naming them but there, there's I think, I don't know, probably 10, 15 protein design startups. I think the, the two things which it sounds like Chai has gone on is like one, all in on product and two, you are not trying to do your own platform if you don't have your own data mode. You know, is that going to like help you win out in the end or is that going to be a, you know, a blocker? I don't, I'm just, I'm just curious about that.
1:19:22Speaker C:Yeah, that's, that's a great question. Yeah, so, so try definitely no plans of like starting a pipeline. Like we take the partnership model pretty seriously and we I just like from a personal stance I love the incentive alignment between like you know, we make the models better, the partners succeed more and just like you know that iterates on itself. Um, so like I think that's like a pretty unique part of Chai is like one, just being able to partner with a lot of people, two, getting like the feedback on the product.
1:19:48Speaker C:So like you know, knowing that it's very real, this is in like like legit big pharma hands and like they're actually running campaigns on this stuff. Um, so I think it's interesting. We really have to be model forward, model focused. Like we need to keep delivering value.
1:20:03Speaker C:So that puts a lot of pressure like on the research team, the product team first of like to serve these things. The research teams always shoot for like better and better versions. The way I think about this is like if you're a bitter, less impelled forward kind of like thinker or company then there kind of comes a certain point where there's a lot to do on like both the model and data side. But I don't think either is exhaustive.
1:20:25Speaker C:It would be stupid to say like we don't need any more data but it'd also be stupid to say like the models are stuck we only can, like, use data to solve these problems. So I think there's like tons of room to grow on both sides. We're taking like, both very seriously. Seriously.
1:20:37Speaker A:And I would also maybe push back on the no data moat premise. Right. That'd be kind of like saying, hey, like, all the enterprises that work with anthropic, like, you're not letting like, anthropic train on their data. So like, you can't like, build models that are good at enterprise workflows.
1:20:49Speaker A:Right. I think, you know, one, we are investing in this. Right. You know, there are ways to turn compute into data and get, get more.
1:20:55Speaker A:And we're, we're doing those. Right. But then also too, okay, what is the kind of data that you're trying to get? Right.
1:21:00Speaker A:And I think what, what is kind of cool about, you know, working so closely and supporting so many of these partners is we get to really learn about, you know, what is like, the stuff that would be helpful in research. Right. And so rather than doing research in a vacuum, you know, based on what would hypothetically be cool, we're able to sort of kind of do informed research based on like, you know, what our, what our partners have just been very organically asking us for help with.
1:21:23Speaker B:I see. Do you, I assume that you aren't allowed to train general models based upon your partner's data. Do you train specific specialized models for like, is there Novartis model in a Pfizer model?
1:21:34Speaker A:Yeah, I mean, like a lot of these, a lot of these deals, you know, and this is all public, right. We are working with them to, you know, train or fine tune a version of our model for them. And I think there's probably like, so much more we can do there over time. My brother started a company called Applied Compute, a great company that kind of doing this thing for, you know, design for lms.
1:21:53Speaker A:Right. And helping enterprises really understand the, the value of their language data and do that for specialized tasks. I think there's a whole world where we could potentially do that for biological data.
1:22:03Speaker D:What is the value there? Like, what is the lift that you get from using their data? I mean, is it just that it's more data or is it more that there's is specialized to a problem?
1:22:12Speaker A:You know, they have, they have a lot of like, scientific, you know, data that they're getting from experiments that can maybe help our models do better in like, particular classes of candidates that are targets that they care about.
1:22:23Speaker C:Yeah. I mean, even something as simple as, like, they might just have some preferred way of Doing things that might not be like native to the Chai model. And they can like, you know, kind of like ask the product team in a sense to just be like, hey, we like, you know, our, our designs have property X. Can you make sure that they have those?
1:22:39Speaker C:So I think like even things as simple as that actually have like a pretty big impact for them.
1:22:46Speaker D:Yeah. So I mean this, this goes along with the, a pet hypothesis that I have that all AI companies and especially bio and scientific ones are actually consulting companies. Pharma I think is particularly the case because you're developing a new drug.
1:23:04Speaker D:It's almost by definition new. Right. So like the existing stuff has to be customized in many cases. Right.
1:23:10Speaker D:Unless you're doing something that's just reiteration of old stuff. But a lot of the big pharma are pushing the boundaries of science.
1:23:17Speaker A:Yeah. I mean certainly like we aim to make the models very general, we aim to make the product very general, we aim to make it powerful. But yeah, I mean there is integration work. Right.
1:23:25Speaker A:With every, with every customer. To answer your question, you do get some defensibility just by doing that. Right. And I think what is, what is nice about building, you know, trusted relationships with these partners is hopefully, you know, if we execute really well over the next, you know, the first year, then they'll continue working with Chai to do more ambitious and more, more drugs past that.
1:23:46Speaker D:I mean, it's going to be hard to switch, right?
1:23:48Speaker A:I hope so. Yeah.
1:23:49Speaker D:Just getting the security review done.
1:23:52Speaker A:Yeah, yeah.
1:23:53Speaker C:Like maybe, maybe. One other interesting point is like if you think of this like on a per token base basis, I don't know if there's another domain where the downstream value of a token is as valuable as it is for pharma. That makes sense. You're thinking about the actual drugs that come out.
1:24:06Speaker C:These can be multi billion dollar assets. The case of GLP1s, I think the two GLP1 drugs combined, maybe a trillion dollar asset.
1:24:14Speaker A:Yeah. I mean up until I think three months ago, right. GLP1's total revenue was more than all of the AI labs put together. I don't think people realize that.
1:24:23Speaker A:I didn't realize that. It was crazy.
1:24:25Speaker B:Right? But yet the market cap way lower. It's crazy how relatively speaking to market.
1:24:29Speaker A:Cap and you know, I didn't realize how much of like a VC business, you know, you know, pharma is in. Right. They're in some sense like taking really ambitious bets. You know, I think one of the things that was really cool, you know, is like if you study the history of Silicon Valley.
1:24:44Speaker A:Right. Like obviously people think of Silicon Valley with software, but you know, in the 80s, one of the, one of the biggest venture outcomes, one of the first ones was, was Genentech. Right. And because it is such a VC model.
1:24:56Speaker A:Right. You get the string of tokens that can then give you so much value downstream.
1:25:01Speaker B:Just, just general shout out to Outposting's blog, blog series about like finance and funding and. Yeah, really fantastic. Yeah, before that I knew a lot of those points, but I did not realize just how deep that rabbit hole went. Yeah, it's.
1:25:16Speaker B:Yeah. I mean it's maybe the biggest single biggest problem in biopharma is actually just the funding model. Model.
1:25:23Speaker C:There's also. Have you, have you heard of Aram's law?
1:25:25Speaker B:Yeah. Oh yeah, yeah, yeah, yeah. More backwards.
1:25:28Speaker C:Yeah, more, More's law backwards. So it's like in like compute, you know, it's kind of scales. So you have like this nice exponential scaling log laner scaling of compute. And you have the exact opposite in pharma.
1:25:39Speaker C:So like the cost of actually making a drug in pharma is kind of like increasing exponentially. So the amount of money put in per drug is growing at kind of like an exponential rate. Which is like. It's pretty interesting to see this, you.
1:25:50Speaker B:Know, which guarantees at some point the marginal return on a new drug development will be negative.
1:25:55Speaker C:Exactly.
1:25:55Speaker B:So unless someone I maybe Chai figures out how to you know, fix this. I think that we might be on the verge of sort of flipping some of these explanations.
1:26:06Speaker D:Bending the S curve.
1:26:09Speaker D:Just to double, maybe belabor the point, but that pharma and VC fundamentally both are optimizing a portfolio.
1:26:17Speaker D:And I think that's the, that's the connection there.
1:26:19Speaker A:Yeah. Thinking of pharma is like sophisticated capital allocators. Right. Where they have this portfolio of targets and they're allocating between them.
1:26:26Speaker A:I think that was a big reframe for me and I think we will just see more of that in the future. Right. And hopefully they can take, you know, in a sense the VC taking riskier bets. Like hopefully pharma can take riskier bets and pursue really, really cool drug targets in the future.
1:26:41Speaker C:That analogy is actually like one the, the kind of like VC type investor ish model. It's like actually how we think a lot about research at Chai is as well our research team is, is relatively small I think definitely compared to like a lot of the, like the isomorphic steep minds. Like our research team is like, you know, in the around 10 people. So like we're, we're a relatively small team, but we kind of think of it as almost like an investing job where like you're investing ideas towards compute in the same sense, you're really just capital allocators in that respect.
1:27:09Speaker A:Yeah, I actually think maybe this is too cute. But I would even make the broader point which is I think every. We kind of think of everyone at CHAI as a bit of a capital allocator. So.
1:27:17Speaker A:So I think one of the things that surprises people is we're pretty small, we're only 30 people. And that's because everyone we hire onto the research team or the engineering team, especially now that they're in some ways very empowered with AI, a lot of it is just allocating their attention into the right ideas and allocating their compute to the right ideas.
1:27:35Speaker D:But this is actually I think a characteristic to some extent of machine learning, AI projects and also science. Whereas if you're building like a API for some B2B SaaS company that's not building foundation models, whatever, your limit is mostly people, right? So you're. The resource you're allocating is almost entirely people.
1:27:57Speaker D:Whereas if you're building hardware, you're building AI models, you're building something scientific, then your constraint is those the resources that are, you know, the bottleneck is, you know, the, the lab, it's the compute, it's other things. And so that you have to really be in that mentality of I have these allocation of I have some shots on goal. How do I allocate those shots?
1:28:20Speaker A:Well, I would say yes and no. So I agree it's a bit more like that. Right. But like let's going back to the example of building an API for, you know, a B2B company, right?
1:28:29Speaker A:That API has incremental cost. You have to support it. It adds complexity to the product. It's another thing you have to go market and sell.
1:28:36Speaker A:Maybe you should actually be allocating that into like a different bet, right? A different thing on your product roadmap that you should be prioritizing instead of the other everything. I think in a world where building things just gets really cheap and increasingly free, the scarce thing is the attention both that you can put into it to keep your product simple and grokkable and that your customer can put into to really understand how to use it. I see it less as a binary thing and more just like we're all kind of as engineers going to be a little bit more allocators of attention,.
1:29:07Speaker D:Which is what executives Are we're all just becoming.
1:29:09Speaker A:Well, I mean, listening to a podcast with Satya at Nadell. Right, Right. He says, you know, Microsoft wants to make everyone a manager of infinite minds. Right.
1:29:17Speaker A:If you, like, really take that to your extreme, like, everyone's going to be an executive. I mean, I certainly feel like an executive. And I talk to Claude every day. Right.
1:29:25Speaker B:Little suite of interns who are all going out and eagerly solving problems you may or may not have actually wanted. But they're solving the problem.
1:29:33Speaker D:Yeah. So we have two typical questions that we asked that we've already kind of asked. One, but I'm going to ask it again, maybe in more direction, correctly, is if you and you can both answer this, if you could remove a bottleneck from your problem space by fiat, what would that be?
1:29:53Speaker C:That's an interesting question. I think one thing that'd be really nice, like, just I'm, like, always in research land, very hard to turn off. For me, it's probably just the validation loop of protein design in general. So just being able to say instantly, like, hey, this thing works, this thing doesn't.
1:30:08Speaker C:There's still a bit of walking around in the dark that you're doing just so, like, you know, you have. You have some ways. And like, I think at chai, we've taken this, like, very seriously, but it's probably along the lines of just, like, validating hypotheses and like, you know, knowing for certain that things work.
1:30:23Speaker D:Yeah, that's unsolved problem, for sure. Unsolved problem. Yeah, yeah. And would be hugely valuable.
1:30:27Speaker C:Hugely valuable. Yeah. Yeah.
1:30:29Speaker A:I'm going to take a much more abstract answer to that, which is actually like, talent obscurity. I think, you know, there's a lot of smart people going and working on LLMs. You know, there's a lot of people that are working and becoming software engineers for. For SaaS.
1:30:40Speaker A:Right. But I think just, like, not that many, like, smart people go and work on bio. You know, I didn't work on bio, like, in high school because I was like, oh, I could, like, pick up my computer and program apps, but if I want to work on bio, I have to, like, go study and get good grades in school and, like, maybe get a PhD or whatever. Right.
1:30:56Speaker A:And, you know, maybe that's one reason for it. I think another reason is, you know, a lot of this stuff is really obscure. Right. Like, we threw around a lot of big words during this podcast.
1:31:05Speaker A:You can't really visualize that things. It's one of the things we care a lot about. At chat is like, how do we make the whole thing feel visual on our website and in the product. And you know, part of the reason we're here is like, you know, I think, you know, more people should realize like, you don't need to like have like a super, super, super specialist bio background to contribute to this, like computationally.
1:31:24Speaker A:And so, you know, I think a lot about like talent flows and like where talent goes in the economy and. Right. You know, in the 90s everyone was flowing to talent and you know, since the 2000s people have been flowing to tech. But, you know, big tech like ate up a lot of the talent, you know, until, you know, a few years ago.
1:31:40Speaker A:And now maybe like LLMs and the big AI labs are eating up a lot of the good talent. But it's like, you know, at the meta level, like, how do you allocate talent better? You know, selfishly, I want more talent going into bio. I mean, we probably want more talent going into manufacturing and physical world things and these other problems that the US has.
1:31:57Speaker A:But yeah, I think communicating that better would be the thing that if I had a megaphone to talk to everyone, I would try to do that.
1:32:05Speaker D:Okay, so, and then that leads to the second question, which is, and maybe the answer is the same, but what is the takeaway, one takeaway that you would like to people to have from the episode?
1:32:18Speaker A:Yeah, I mean, I think, you know, biology has been this somewhat obscure feeling field where you're stumbling around in the dark, you don't know what you're looking at. You are dealing with non determinism in your experiments. You're having to do a very long and iterative trial and error loop across a very, very long amount of time. And at some point you're crossing that threshold of what you can do computationally.
1:32:44Speaker A:When you can get folding models down to being within an angstrom, where you can get design models to give you hit rates north of 50%, where now you can put them in a 96 well plate and actually have like 48 interesting binders. You start to get to the point where now you can declaratively precision engineer what you want rather than betting on nature or trial and error to get you there. And I think that, look, we had the same thing happen in software, where you can write code and you can deterministically get an outcome. Or in electrical engineering, where instead of your schematic being drawn out, you can put it in cadence design system systems and get it made in software or CAD for mechanical engineering where you can sort of precision engineer your part and get it printed or manufactured.
1:33:34Speaker A:The same thing is happening in bio and it's happening very quickly and that really opens the door for a lot of really interesting people. Or maybe it wasn't as inscrutable or accessible before. Right. Like software engineers like myself, researchers like Matt at, you know obviously we're still going to want the specialists but you know the, the, the, the generalists can often really accelerate the, the precision engineering happening in the domain.
1:33:58Speaker C:Yeah, I think for, for me like the base takeaway is that the field is actually working and like, like not only does it have commercial traction but like the research is like actually showing signs of life. Like it's not even just showing signs of life. Like the signs of life have been showed. We're actually in a place where like the models work, they're delivering value, value and like there's still tons of really interesting research problems to solve.
1:34:19Speaker C:So I think there's a lot more low hanging fruit in this field than there would be in other fields. And I think the amount of impact that you can have especially like as a researcher is just like unmatched in this, in this field for us. We're all very mission driven. But even if you're not like it's a lot of fun puzzles to solve.
1:34:34Speaker C:Like there, there's like this kind of 3D geometry angle, there's like, if you like diffusion models, there's like a million problems to solve in that regard. We have this LLM looking trunk in like Chai one. There's just so much of like core machine learning is touched by these problems. We're still, although we've made a ton of progress, there's still a lot to be done and I think it's just like one of the most interesting fields to be working in which like while also having some of the largest impact on just like humanity.
1:35:00Speaker D:Thank you so much for having us for making a long journey.
1:35:04Speaker C:22 Minute walk.
1:35:07Speaker D:And you know we look forward to tracking Chai's progress.
1:35:09Speaker C:Awesome.
1:35:10Speaker B:Thank you guys.
1:35:10Speaker D:Thank you very much.
