00:02:38 · Josh Meier, Sachin, Adam Miller
[00:02:38] S0: Yeah, yeah, that's for sure. I think if you ask Paul on our team, asked him like, what's your favorite part about Chai? And he's like, the results. And I love that. I think we're, we're very, very mission driven, right? And a lot of that does come down to the results because if the models don't work well enough, then we can't make the best medicines possible. So that's why, you know, it all kind of comes back to how do we make the best technology? How do we make that technology accessible and work well for our customers? So it's a very exciting time to see us all come together. [00:03:07] S1: Maybe a funny place to start is we were chatting to Jack and he was telling me about how you guys met at Harvard and you're in the same classes. We asked him this question, so let's ask you, who's a better software engineer out of you two? [00:03:17] S0: I mean, Jack by far. I like so much for him like every day. [00:03:20] S2: Yeah, I love the humility. [00:03:23] S0: It's the truth. I would love [00:03:25] S2: to start sort of way back and just understand that when you were growing up, do you think you had signs in your upbringing, maybe in high school, that one day you'd be running a company like Chai of this size and trying to take on a mission like this?
00:03:25 · Adam Miller, Josh Meier
[00:03:25] S2: to start sort of way back and just understand that when you were growing up, do you think you had signs in your upbringing, maybe in high school, that one day you'd be running a company like Chai of this size and trying to take on a mission like this? [00:03:39] S0: I don't think I ever thought I would necessarily build a company. I'm not surprised when I ended up in this space growing up. I grew up as coder, right, as a young kid, but I come from a family of doctors, so I always wanted to do something in medicine, but I just loved programming, so I was like, programming seems like way more fun than just like, you know, showing up and treating patients every day, it seemed more repetitive to me, and of course it's not, and like doctors do amazing things, right, and they're so important for actually bringing these medicines to patients, but I always loved the scale that you got from programming, and that was what was so exciting about the process of making medicines, like one medicine, we're in the world of AI now, tokens, like the tokens that go into these medicines are so valuable, right? And they can be produced again and again and help so many people around the world. And I think if you have the skills to contribute to this in some way, it's almost like a moral obligation. Like this stuff is so important for humanity and society, and if you have the opportunity to work on it, you're in AI right now and you can literally turn like AI, you can turn GPUs into drugs, it's like a no brainer to work on something like that.
00:04:42 · Sachin, Josh Meier, Adam Miller
[00:04:42] S1: What's the biggest, like, motivator for you in terms of working on this mission? Like Josh told, sorry, we were told before that Alzheimer's is something that we think in the foreseeable future we could cure. Is an example like that of something that you'd love to just go after with the technology? [00:05:00] S0: I think for me, it's actually more broader scale. Like, it feels right to me that this is the way that most future medicines are going to be made. If you just think about it from first principles, if we are able to actually learn, in a sense, the laws of how atoms interact with one another, how pathways work, right, how medicines can interact with a cell and eventually with an organism. It just makes a lot of sense that that's how most future drugs would be made, and that's what I'm most excited about, just the broad scale that we get from that as opposed to any one disease. [00:05:31] S2: And you started out as one of your first jobs working for OpenAI, so how did you go from training models to training molecules to what you're doing now?:
00:05:31 · Adam Miller, Josh Meier
[00:05:31] S2: And you started out as one of your first jobs working for OpenAI, so how did you go from training models to training molecules to what you're doing now?: [00:05:39] S0: Yeah, so I studied chemistry actually in school, and I was a programmer since I was a kid. I did a lot of computer science in college too, But, so at OpenAI, we were training some of the early language models, right, GPT-one, GPT-two was happening then, and it occurred to me that if language models worked for natural language, French and German and English, why wouldn't they work for the real natural language, like DNA and proteins? And that's really how I got into it. I kind of just got obsessed with that idea at that point. Even back then, you know, like the business model around AGI was not very clear, right? It's like you had these, very mediocre chatbots and you had to take a ton of generations from the models to get them to do something interesting. But in biology actually, if you could take a 100 generations and like just, well, just one of them was good, that's actually incredible. People are used to screening millions or billions of molecules just to find like an initial starting point where you go make it better. So it felt to me actually that like the ingredients in AI were actually set up perfectly to have some impact in biology.
00:05:39 · Josh Meier, Sachin
[00:05:39] S0: Yeah, so I studied chemistry actually in school, and I was a programmer since I was a kid. I did a lot of computer science in college too, But, so at OpenAI, we were training some of the early language models, right, GPT-one, GPT-two was happening then, and it occurred to me that if language models worked for natural language, French and German and English, why wouldn't they work for the real natural language, like DNA and proteins? And that's really how I got into it. I kind of just got obsessed with that idea at that point. Even back then, you know, like the business model around AGI was not very clear, right? It's like you had these, very mediocre chatbots and you had to take a ton of generations from the models to get them to do something interesting. But in biology actually, if you could take a 100 generations and like just, well, just one of them was good, that's actually incredible. People are used to screening millions or billions of molecules just to find like an initial starting point where you go make it better. So it felt to me actually that like the ingredients in AI were actually set up perfectly to have some impact in biology. [00:06:36] S1: And this actually isn't that much of a commentate. Like traditional pharma will say that AI falls apart in biology related problems. What's your kind of response to that? How did you know that there was something here where you could apply some of things you learned at OpenAI to solving this problem?
00:06:52 · Josh Meier
[00:06:52] S0: Well, I think if you asked me that question a couple months ago, I probably would have said they're correct, right? Like on the whole, this tech has been very promising, but it hadn't really reached the inflection point for it to be useful because we have to keep in mind, like, discovering a drug is very, very hard, right? Like, we can't see things at the atomic level ourselves. The search space is massive. There's more possible molecules than there are atoms in the universe. So there are, there's so much to search through and we have very primitive tools for getting there. So as a result, have spent decades building very, very good tools to screen through tons of molecular matter. When I say tons like millions to billions, right, it's still a small fraction of the search base, it's actually AI has quite a high bar to go against. You need to do better than what people are doing in the lab today. I think it turns out that, you know, where we're at right now, it's not necessarily the highest bar anymore, right, because AI has these scaling laws that you don't have in the lab, but it took a very long time for us to get to this point, and it was really only in the last couple of months with Chai two and now Chai three, that we've actually managed, I think, to make a dent on this space.