Skip to content
a16za16z

Mark Zuckerberg & Priscilla Chan: How AI Will Cure All Disease

Priscilla Chan and Mark Zuckerberg join a16z’s Ben Horowitz, Erik Torenberg, and Vineeta Agarwala to share how the Chan Zuckerberg Initiative is building the computational tools that will accelerate the cure, prevention, and management of all disease by century's end. They explain why basic science needs $100 million-scale projects that traditional NIH grants can't fund, how their Cell Atlas became biology's missing periodic table with millions of cells catalogued in open-source format, and why their new virtual cell models will let scientists test high-risk hypotheses in silico before investing in expensive wet lab work. Plus: the organizational shift unifying the Biohub under AI leadership, what happens when biologists and engineers sit side-by-side, and why modern biology labs are expanding compute instead of square footage. Timestamps 00:00 Introduction 03:42 Building tools to accelerate scientific discovery 05:26 The credible path to funding basic science 07:03 Biohub = Frontier Biology + Frontier AI 08:58 Challenges building on a 10-15 year timeline 09:39 How CZI chooses what to work on 11:17 Making sense of science with LLMs 11:32 Measuring success in the therapeutic realm 13:32 "Most diseases should be thought of as rare diseases” 15:39 Inspiration: building a periodic table for biology 19:27 Why virtual cells? 21:17 The Biohub Master Plan 21:51 How virtual cell models allow more risk taking 28:15 Bringing CZI & Biohub together 30:32 Why Biohub matters 33:36 The importance of interface design in democratizing scientific discovery 35:34 How Biohub encourages cross-functional collaboration 40:38 Looking ahead: the broader impact of AI on biotech Stay Updated: If you enjoyed this episode, be sure to like, subscribe, and share with your friends! Find a16z on X: https://x.com/a16z Find a16z on LinkedIn: https://www.linkedin.com/company/a16z Listen to the a16z Podcast on Spotify: https://open.spotify.com/show/5bC65RDvs3oxnLyqqvkUYX Listen to the a16z Podcast on Apple Podcasts: https://podcasts.apple.com/us/podcast/a16z-podcast/id842818711 Follow our host: https://x.com/eriktorenberg Please note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details please see a16z.com/disclosures.

Mark ZuckerbergguestPriscilla ChanguestErik TorenberghostVineeta Agarwalahost
Nov 6, 202544mWatch on YouTube ↗

EVERY SPOKEN WORD

  1. 0:003:42

    Introduction

    1. MZ

      This is a, a, a space that, I mean, that there's just gonna be a huge amount of leverage with AI. It still seems like there could be a lot more effort in this space around building tools, and it's kind of this crazy thing that we're, you know, here in, you know, 2025, and there's not the kind of periodic table of elements equivalent for biology. We think that this is, like, probably one of the most important sets of tools that you need to build.

    2. PC

      When we first set out that-- the goal to cure and prevent disease by the end of the century, people-- Like, honestly, most scientists couldn't look at us with a straight face.

    3. MZ

      [laughs]

    4. PC

      And because-

    5. MZ

      They're like, "You're crazy."

    6. PC

      Yes.

    7. MZ

      [laughs]

    8. PC

      And it was true because if you just decided to spend the money funding the next best grant for every single lab in the country, like, you-- there's no pathway to that being true.

    9. MZ

      The biology folks, I think, looked at it as if it were crazy ambitious, and then the AI folks are like, "Well, that's kind of boring. That's just automatically gonna happen."

    10. ET

      [laughs]

    11. MZ

      So, I mean, it's like, okay, there's something in between there that needs to be bridged.

    12. ET

      [upbeat music] Mark, Priscilla, welcome to the a16z Podcast.

    13. PC

      Thanks for having us.

    14. MZ

      Yeah. Great to be here. Excited.

    15. ET

      All right. Excited to have you. You're doing exciting stuff. Yeah. Well, to, to that end, almost a decade ago, you guys started the Chan Zuckerberg Initiative with the mission and intent to cure, prevent, manage all disease by the end of this century. There's a lot of missions that you guys could have poured your time and resources into. Why don't we talk about-- Take us behind the conversations of why you guys picked this one. Maybe Priscilla, why don't we start with, with you and you-your, your side of the story?

    16. PC

      It always surprises people when I talk about how we work in basic science research. Um, I trained as a pediatrician, and people always think like, "Oh, it must be about medicine." And for me, it wa-- Uh, you know, I went into medicine because I wanted to improve people's lives. I wanted to make a difference. I wanted to be able to help others. And I think training as a pediatrician at UCSF, I met a lot of patients and frankly, like, little kids and families for which, like, we just had no idea what, what the problem was. And they might have, like, a specific gene that they could name if they were lucky, um, or they could be grouped into a bunch of other diseases, and there'd be a general sort of PDF they'd print out, like, "This is what we know." And then it was my job as an intern or resident to try to translate, like, uh, like, a few lines of information to how we were g- supposed to take care of the patient. And for me, that's when I really, like, realized the power of basic science and how we need to work on basic science to advance the forefront of what's possible. And without that, there's sort of-- It-- I think of it as the pipeline of hope.

    17. ET

      Yeah.

    18. PC

      Hmm.

    19. ET

      And why did you think, um, you could cure all disease? 'Cause that's, like, a very, like, aggressive goal.

    20. PC

      Um, do you wanna, do you wanna answer that one?

    21. MZ

      Yeah. Well, well, I mean, we're not gonna cure all diseases-

    22. ET

      Yeah

    23. MZ

      ... to be clear. I mean, the, the strategy is to help scientists and the scientific community cure all diseases.

    24. ET

      Yeah.

    25. MZ

      So the strategy is really one of accelerating the pace of basic science. And the theory that we had was, if you look at the history of science, most major breakthroughs are basically preceded by the invention of a new tool to observe phenomena in a new way, right?

    26. ET

      Right.

    27. MZ

      So think about things like the mi-microscope, right?

    28. ET

      Right. Right.

    29. MZ

      Being able to, you know, observe bacteria or-

    30. ET

      Sure

  2. 3:425:26

    Building tools to accelerate scientific discovery

    1. MZ

      so our, our whole approach on this is basically let's help build tools that will accelerate the pace of the whole field. And I think that that-- There's a niche that I think fits that, because if you look at how funding works in science, you know, the vast majority of funding comes from the government and NIH grants. It's parceled out into these relatively small grants that allow individual investigators to investigate usually pretty near-term things.

    2. ET

      Right.

    3. MZ

      Um, and the development of these kind of new types of tools, whether it's imaging or building now a lot of AI things like virtual cell models, um, are longer term, oftentimes more expensive to develop. So think about, like-

    4. ET

      Right

    5. MZ

      ... on the order of a hundred, you know, m-maybe, you know, a hundred million to a billion dollars over a, um, over a ten to fifteen year period, and then you, you try to unlock those tools and give them to the scientific community to accelerate the pace. So that's, that's kind of the, the theory.

    6. ET

      Right. And, and there-- It seems like there's also something that, that, that-- is you don't really get credit for the tools in a lot of ways. I mean, we've been noted, we have companies that, uh, use your tools and very happy about it. But, um, you know, I didn't even know that that was the case, and so-

    7. MZ

      That's why it's philanthropy.

    8. ET

      Yeah, well, it is.

    9. MZ

      [laughs]

    10. ET

      But most people do philanthropy to get credit too. I mean, you know, like, that's a, you know, that's kind of a part of it. So how did you-- I, I guess, did you think about that, or were you just like, "No, like, this is gonna work, and if it works, that's all we need"?

    11. PC

      We're s-super focused on-

    12. ET

      Yeah

    13. PC

      ... like, actually making every scientist better, and, and beyond science, like startups, startup founders, because I-- the point is, we can't do this alone.

  3. 5:267:03

    The credible path to funding basic science

    1. PC

      And when we first set out that-- the goal to cure and prevent disease by the end of the century, people-- Like, honestly, most scientists couldn't look at us with a straight face.

    2. MZ

      [laughs]

    3. PC

      And because-

    4. MZ

      They're like, "You're crazy."

    5. PC

      Yes.

    6. MZ

      [laughs]

    7. PC

      And it was true because if you just decided to spend the money funding the next best grant for every single lab in the country, like, you-- there's no pathway to that being true. But if you forced people to really think about this and like, "Okay, what is the most credible pathway to doing this, and what are the barriers to that credible pathway?" Then we sort of got somewhere, right? They were like, "Well, like, there's no shared tools," or like, "We don't have-- We're not working on big projects and building the right datasets." And we're like-Okay, well then we can start-

    8. VA

      Let's do that

    9. PC

      ... doing something about that. Um, and so that's where the idea of building shared tools, 'cause no one right now in the scientific-

    10. VA

      Oh, that's so interesting. So basically you're like, "We're gonna cure all disease," and they're like-

    11. PC

      Can't

    12. VA

      ... "Yeah, c-can't be done." Why can't it be done? Well, 'cause we don't have the tools. Okay. [laughs] That's pretty, that's a pretty cool sequence.

    13. MZ

      Yeah, I mean, there's also this funny thing where the, the biology folks I think looked at it as if it were crazy ambitious, and then the AI folks are like, "Well, that's kind of boring. That's just automatically gonna happen."

    14. VA

      [laughs] Yeah.

    15. MZ

      So I know, that's like, okay, there's something in between there that needs to be bridged.

    16. PC

      Yeah.

    17. MZ

      And if you can like kind of use the, the kind of modern AI tools in order to build the types of tools that biologists need. So that's a big part of how we think about our work is, um-

    18. VA

      AI's got to be the most overestimated and underestimated technology [laughs] ever.

    19. PC

      [laughs]

    20. VA

      Like simultaneously. It's so weird.

    21. MZ

      I, I mean, yeah, well probably like the internet early on.

  4. 7:038:58

    Biohub = Frontier Biology + Frontier AI

    1. VA

      Yeah. Yeah, yeah.

    2. MZ

      But, but we kind of think about ourselves and the work that we're doing at the Biohub as frontier biology paired with frontier AI.

    3. VA

      Yeah.

    4. MZ

      Right? So there's-

    5. VA

      Oh, right

    6. MZ

      ... there are labs that do frontier AI that, uh, basically, you know, are building the most advanced models. Um, and then there are lots of biological research organizations that, that effectively do very leading edge-

    7. VA

      Mm-hmm

    8. MZ

      ... research to build, um, you know, to either discover new data sets or, or, or-

    9. VA

      Right

    10. MZ

      ... looking to certain challenges. But so far there hasn't been anyone who's tried to do both of those at once. And when you look at, I mean, even something like AlphaFold, which is amazing-

    11. VA

      Mm-hmm

    12. MZ

      ... right? It's, it was built off of this data set-

    13. VA

      Right

    14. MZ

      ... that was a public data set that had been produced d- decades ago, right? And, um, what, what I think you have the opportunity to do if you do both of those together is produce specific data sets for the purpose of training AI models to build virtual cells that can do specific things.

    15. VA

      Yeah. Right. Right. Right.

    16. MZ

      So I think that that's like a, a pretty interesting zone to be in.

    17. VA

      Yeah.

    18. MZ

      And of all the things that, that we've, uh, that we've worked on, you know, actually, when, when we started CZI, we, we kind of actually focused on a number of areas, and what we found is just that the science research has had by far the biggest returns, so we've just doubled down on it over and over and over until now we're at the point that, you know, we're 10 years in, and Biohub is really the, like, main focus of-

    19. VA

      Mm

    20. MZ

      ... of our, of our philanthropy at this point. Um, but yeah, I mean, that's kind of, that's basically the focus.

    21. VA

      Well, I think if you-- maybe you're not giving yourselves enough credit 'cause you're sort of saying, "Well, there's bite-sized science. We don't wanna do that. There's century s-scale science [laughs] , and that seemed like a long time horizon, but achievable, ambitious." But you've actually identified, you know, which I think is really fantastic, grand scientific challenges that are right in between. They're 10 to 15-year horizons, at least per-

    22. MZ

      Mm-hmm

    23. VA

      ... kind of the way you communicate about them-

    24. PC

      Yep, that's right

    25. VA

      ... and the way you energize-

    26. MZ

      Yeah

    27. VA

      ... the scientific

  5. 8:589:39

    Challenges building on a 10-15 year timeline

    1. VA

      community about them. 10 to 15 is kind of an interesting time horizon, sort of like similar to the time horizon of a venture-backed company, similar to the time horizon on, on which a team can work together for that period of time. I think that's-- how did you get to that number, and then how are you thinking about the challenges that you take on in each 10 to 15-year wave? Because that's concrete, achievable, you know, you build a lot of credibility around it the way that you've announced those challenges.

    2. PC

      Well, I'm curious how you guys think about it, but for us, when we looked at the grand challenges for, on the 10 to 15 ti- year time horizon, it needs to be like, when you look at it, y- you're like, "I see a path."

  6. 9:3911:17

    How CZI chooses what to work on

    1. VA

      Right.

    2. MZ

      Mm-hmm.

    3. PC

      Not everything needs to be solved for us to take it on. In fact, if everything's solved, then that feels like that should just go.

    4. VA

      Then it wasn't ambitious enough.

    5. PC

      Yeah.

    6. VA

      [laughs]

    7. PC

      Like you, like we, we have, we have some risk appetite, right?

    8. VA

      Yeah.

    9. PC

      So we want things where we're like, there's a credible pathway, someone, uh, who is at the helm who can do this, and there's enough ambiguity where we feel like we could take on that risk, and if we do it, like the, the returns could be higher than even expected. And the way we modeled that from, you know, in the Biohubs is we, we have three Biohubs. We have o- one in San Francisco, one in Chicago, one in New York. The one in New York works on cell engineering. You know, can we engineer cells to go in and detect signals, read it out, or to take certain actions? In Chicago, we're building tissues and looking at, uh, tissue c- ce- cell-cell communications within tissues. And then in San Francisco, we're looking at deep in- imaging and, uh, transcriptomics. And that work, the locations are not by accident. We also look at the partner universities because we have folks who come to the Biohubs to do this work, collaborative, interdisciplinary, um, and sort of unconstrained by the traditional lab. But we also build off of the labs at these academic institutes that support the work. And so, uh, that's how we sort of choose the grand challenge and, um, and the locations. And then the sort of layering or sh- and the, uh, l- large language models and AI coming into the picture has been so interesting

  7. 11:1711:32

    Making sense of science with LLMs

    1. PC

      because we were already building tools to measure interesting data, building the data sets, but we didn't really know what to do with them yet. Um, and large language models coming onto the scene, we're like, "Wow, we can make sense of all of this now."

  8. 11:3213:32

    Measuring success in the therapeutic realm

    1. VA

      I'm curious what you view success as f- in the therapeutic realm. So, you know, we think a lot about understanding biology, and sometimes we bet on startups that wanna unlock completely new biological areas, diseases where we don't know what's going wrong. And then there's another group of folks who kind of say, "Hey, okay, now that we understand what's going wrong, let's fix it."

    2. PC

      Mm-hmm.

    3. VA

      Um, let's come in with a drug. Let's come in with a new type of chemistry, a new type of antibody. How do you-- what do you think success for the CZ Biohub looks like 10, 20, 50 years from now?In terms of the new medicines that you've enabled

    4. PC

      We want there to be like an explosion of a community who are building these, um, just the new wave of what it means to be deploying precision medicine. Like w-- like I think for rare diseases and common diseases alike, you're really talking about individual biology that we sort of lump together. Um, and, uh, they-- and we often don't know how it happens, right? We know that you have this mutation or the worst nightmare is you have a variant of unknown significance. What does that even mean?

    5. VA

      The horrible VUS.

    6. PC

      Yes, horrible. And you're like, you tell someone you kind of know something, but we don't know what it means. But if you look at the way we've been able to look at variants and look at single cell transcriptomics, we're starting to be able to say, "Okay, this variant actually impacts this set of downstream cells." And then we start looking at the proteins that get expressed and w- how it looks similar or different to what a healthy cell would look like. Then you can start targeting, okay, like let's look at that as a target, and you both know the specificity of the target you want to build based on the bil-- the ability to connect mutation to protein expression, as well as to be able to predict off, uh, target effects. What are the side effects? Because you also know where else that drug will s- be able to interact with the body. And, and so those are rare,

  9. 13:3215:39

    "Most diseases should be thought of as rare diseases”

    1. PC

      like and, and but I really think most diseases should be thought of as rare diseases because each one of our biology is different.

    2. VA

      Yeah.

    3. PC

      And right now we just get lumped, right? We get lumped based on age, demographics, ancestry, if we're lucky, uh, to have that level of understanding. But truly, each one of our biology is different, and say like if you look at hypertension or depression, like we kind of just go by trial and error and saying like, "Let's just try that drug and see what happens." But what should really happen is being able to precisely and accurately and quickly treat people by looking at individuals' biology. We want to enable the basic science, and we would be thrilled if people picked up the models that we build to be able to build the diagnostics, the therapeutics that need to come.

    4. VA

      You've built amazing data sets, I have to say. Like, I mean, you may not hear the feedback from the startup community and the pharma community and the R&D community, but it's there because you've committed to open source. And so people may not be-- they may not all be writing papers, but they are using those tools. Um, there's a startup in our portfolio working on idiopathic pulmonary fibrosis. The name tells you how vexing the disease is. It's idiopathic.

    5. PC

      Mm-hmm.

    6. VA

      We don't know [laughs] why it happens, but IPF is named that way. And so, you know, he was telling me that he used your Cell by Gene atlases-

    7. PC

      Mm-hmm

    8. VA

      ... to look at millions of single cells in patients with disease, without disease, try to pinpoint the fibroblasts, double-click on the fibroblasts and their gene expression-

    9. PC

      It's incredible

    10. VA

      ... and try to, you know-

    11. MZ

      Yeah

    12. VA

      ... use that to inform, hey, where could I go after a new drug target in this disease that's fundamentally a strange clump of idiopathic, you know, idiopathic, um, origin. So, um, I think there's a huge, there's a huge group of innovators who are, who love the tools, the visualizations, the query systems, and really the software approach-

    13. PC

      Mm-hmm

    14. VA

      ... that you've built to making that data incredibly accessible. So thank you.

    15. PC

      Cell by Gene is like almost an accident, though.

    16. VA

      [laughs]

    17. MZ

      [laughs]

    18. PC

      Um-

    19. VA

      Tell us more.

    20. MZ

      That's it, yeah.

    21. PC

      So do you want to share a little bit about Cell by Gene, or do you want me to start?

    22. MZ

      Well, I mean, I don't know which part you want to get into, but I mean, but the

  10. 15:3919:27

    Inspiration: building a periodic table for biology

    1. MZ

      Cell Atlas work overall, I mean, it's kind of this crazy thing that we're, you know, here in, you know, 2025, and there's not the kind of periodic table of elements equivalent for biology, right? So that was sort of a lot of the inspiration of it was, all right, how do we both, through work that we're gonna do in the Biohub and through other grants, um, be able to pull together and standardize a format where you can have all this data? And when we were starting off, we didn't even necessarily have in mind that we were going to use that to build virtual cell models.

    2. PC

      Yeah.

    3. MZ

      I think that that's sort of just come into focus as the AI work has advanced, but that's a very exciting thing. We should definitely spend a bunch of time on the virtual cell models, but I'm not sure what you wanted to get into on the Cell Atlas.

    4. PC

      Well, the single cells work is, was one of our first RFAs 10 years ago we started.

    5. MZ

      Mm-hmm.

    6. PC

      And we were like, "Okay, we think this is possible." We actually funded the methodology for it to, to standardize how it was gonna be done. So that was 10 years ago. And we then were-- we seeded a few labs to start building out that data set, but we were like, there are like millions or billions of different cell types and different permutations. Like, how are we going to do this? And, um, especially with, like, a burgeoning technique. And so we ended up, um, seeding a few groups, and they started doing work, and then they told us they had a problem. There was a, uh, there was a, a bottleneck in their workflow because they couldn't annotate the data fast enough. Um, and so we built, Cell by Gene was an annotation tool. That's the original source of this.

    7. MZ

      Huh.

    8. PC

      So we built the annotation tool to make it easy for people to, who are doing single cell science to be able to annotate the data. And then we put, we put the data that we collected publicly so people could share. But because everyone started using the same annotation tool, everyone was standardized then on the same data formats.

    9. VA

      Hmm.

    10. PC

      And then there started being a f- a community around the tool, and they wanted to share back and build the atlas. So now after 10 years, there are millions of cells that have been built into this, uh, shared resource for the entire scientific community. We only funded about seventy-five percent of it. Sorry, that's wrong. We've only funded twenty-five percent of it. Seventy-five percent came from the broader community saying, "This is useful, and there's an easy way for us to standardize and build these together."

    11. VA

      The default, you have the same metadata.

    12. MZ

      Yeah.

    13. PC

      That's right.

    14. VA

      Formatting format.

    15. MZ

      It's like an interesting, it's like what you'd call a network effect, right? [laughs]

    16. VA

      Yeah.

    17. MZ

      Exactly. [laughs]

    18. VA

      Yeah, I was gonna say, it sounds like the internet. Yeah, yeah. Like come for the annotations, stay for the-

    19. MZ

      Yeah.

    20. VA

      Stay for the virtual cell models.

    21. PC

      Mm-hmm.

    22. MZ

      Well, it was very important when we were getting started with the work to have everyone who was doing it have a consistent format, so that way it could be used and portable.

    23. VA

      Yeah.

    24. MZ

      And then once that kind of took off as, as the way that it would get done, then other people just found it valuable too.

    25. VA

      Yeah, and even relative to prior data bases like Geo and, and whatnot, they're just simply not-

    26. MZ

      Yeah

    27. VA

      ... as standardized or QC.

    28. PC

      Yeah.

    29. MZ

      Yeah.

    30. VA

      Yeah.

  11. 19:2721:17

    Why virtual cells?

    1. MZ

      But we think that this is like probably one of the most important sets of tools that you need to build. Um, and it's not a single thing, right? So there's different angles to, to come at this from. The Cell Atlas data is helpful for understanding things on a cellular level. Um, one of the, the kind of most important things that we're doing right now, the, the, um-- There's this, this great company, Evolutionary Scale, who actually had a bunch of researchers who'd formerly worked at Meta on protein folding models, um, is joining a, a Biohub, and, and Alex Reeves, the, the, uh, leader of it, is actually gonna be the, the kind of head of the whole science program, which is actually kind of interesting-

    2. ET

      Yeah.

    3. MZ

      When you think about it, where it's like you have AI and biology coming together, and really it's like an AI person who understands biology is running it rather than a biologist-

    4. ET

      Right. Yeah

    5. MZ

      ...who has some understanding of AI, I think just kind of speaks a little bit to where we think the, the relative, um, weight of these things is. But I mean, we basically view, you know, like Priscilla was saying with the different Biohubs, and then New York doing cellular engineering will basically make it so that you can have cells that can record different things that are going on around the body and, and share that data, and then you can build that into models. The Chicago Biohub being able to record inflammation, um, and, and basically study that in order to kind of help understand, um, like that-that's a, that's a different data set. We have the Imaging Institute, which is w-we just trained our, our first set of models around that, which are the first like spatial models around understanding like the way that, that kind of cells look in different states. And eventually, just like you have this analogy on the, um, kind of the industry side around language models where you have different capabilities, and then over time you train them into models and it gets more and more general.

    6. ET

      Right.

    7. MZ

      That's kind of the idea here. So we'll-

  12. 21:1721:51

    The Biohub Master Plan

    1. ET

      Huh.

    2. MZ

      We'll, we'll build the Biohubs around grand biological challenges. The Biohubs will build tools that will generate novel data sets. We will build models based on those and then eventually combine the models into an increasingly general view-

    3. ET

      Mm-hmm

    4. MZ

      ...of a virtual cell that will be useful, um, both for scientists and hopefully startups and companies that are working on finding drugs, which is not our part of the whole thing-

    5. ET

      Mm-hmm

    6. MZ

      ...but, but I think is obviously a really important part of what needs to happen.

    7. PC

      Yeah. And you know, you guys think about risk all the time in terms of when you make investments. Like, I think

  13. 21:5128:15

    How virtual cell models allow more risk taking

    1. PC

      the promise of being able to do virtual biology using a virtual cell model is you can actually take on riskier ideas. Right now-

    2. MZ

      Yeah

    3. PC

      ...like grant funding can be hard to come by, and the wet lab work is expensive and slow, and it's not just, you know, money, it's also time. And so you have to choose something that you think is gonna have some likelihood of success to keep your lab career going. And so it naturally lends people to take on like some risk, but not a lot of risk 'cause they need to make sure that they are hitting like a certain percentage of the time to make tenure or publish or whatever they need to do. But if you had a virtual cell model where you could simulate really high quality biology, you could actually then start testing and tinkering on the computational side and like ask riskier questions, things that would've been expensive and ti- costly in terms of time and resources to do in the lab, and actually see if there is promise doing the experiments in silico before you make the time and money investment in the wet lab.

    4. ET

      Do you think of it kind of like a model organism?

    5. MZ

      Yeah.

    6. ET

      Like it's the new fruit fly?

    7. MZ

      Yeah. [laughs]

    8. PC

      [laughs]

    9. ET

      I was gonna ask, given the complexity of a cell, like how close, um, like how accurate do you think you'll get the model to? I mean, just assuming, I mean, maybe you get it to like a perfectly accurate representation of a cell, but like how accurate to be useful would the virtual cell have to be?

    10. PC

      I think it will obviously iterate-

    11. ET

      Yeah

    12. PC

      ...and get better and better because right now we, we-- like right now we're still just talking about, uh, transcriptomics. We're expanding into different ways of looking at the cell. But y-you get more and more accuracy and-- but I don't think it needs to be a hundred percent accurate to be useful.

    13. ET

      Mm-hmm. Yeah.

    14. PC

      Because you just want to be able to de-risk the idea on the front end a little bit. Um, and the more and more you de-risk it, the, the more efficient it gets obviously.

    15. ET

      Right.

    16. PC

      But it'll be useful to, if you even get directional signal. And yes, I do, we do think about it like as a, a model organism, but in a way that's like has fidelity to the human body. Like, you know, like I don't want us-

    17. ET

      All models are wrong, some are useful.

    18. PC

      Yeah. [laughs]

    19. MZ

      Yeah.

    20. PC

      Yes.

    21. ET

      This hopefully-

    22. MZ

      Yeah

    23. ET

      ...has, has utility on certain axes.

    24. MZ

      Exactly. And just like the language models, you build in specific capabilities.

    25. ET

      Yeah.

    26. MZ

      So it's not-- So for example, you know, one of the models that, uh, we're, we're publishing is, is variant former, right? It basically, you know, makes it so that, um, it's trained on a bunch of effectively pairs. If you, you have a cell, you apply CRISPR to it in a place, you see what comes out at the other side. So it's, it basically is able to make that kind of a prediction.

    27. ET

      Mm.

    28. MZ

      Like, okay, if you have this-

    29. ET

      Edit

    30. MZ

      ...edit that you're doing to, to a cell, what is likely going to happen? Um, another one of the models is, it's this diffusion model. Basically, you can describe a type of cell that you would like it to simulate, and it will just produce a kind of synthetic model of, of, of the cell. Um-A-again, I mean, it's kind of interesting because to Priscilla's point before about how everyone is different and, and like, and different cells have, have kind of, um, you know, so you, you wanna be able to simulate these kind of rare configurations, um, having at least a synthetic version of what that could look like is interesting, and then you can test against that. The cryo model I think is interesting 'cause it's spatial.

  14. 28:1530:32

    Bringing CZI & Biohub together

    1. PC

      uh, the big news is, uh, thinking about how we are going to be coming together as one team. Um, and, you know, in the past, we have done, we've run BioHubs, and we've done built software, we've done some AI research. But all of it has been really thinking about, has been a little bit decentralized. But now, under Alex's leadership, we are going to come together as the Biohub, a, uh, an, an operating philanthropy where we are doing the science, um, in service of a singular goal together, and how do we actually advance the state of biology and research, um, at the intersection of AI and biology.

    2. VA

      Amazing.

    3. PC

      Yeah.

    4. VA

      Alex is amazing, so [inaudible] .

    5. MZ

      Yeah, no, he's great. And then, and then the other thing is the, the piece that, that I mentioned earlier, which is just, yeah, I mean, CZI has focused on a number of different things. We've really just found over time that we, we feel like we've been able to make the biggest difference in science, so we've just kept on doubling down on it.

    6. ET

      Mm.

    7. MZ

      And we're gonna continue doing work in education. We're gonna continue supporting local communities and, and in those different pieces. But going forward, the Biohub is really going to be the main thrust of our philanthropy.

    8. ET

      Mm.

    9. MZ

      And we're very excited about that 'cause I think that this is-- There, there has been, you know, when we started, the mission to see if we could help the scientific community cure and prevent diseases by the end of the century. I do think with the advances in AI, that should be possible to do significantly sooner.

    10. ET

      Yeah.

    11. MZ

      And that is a very worthy and important and very exciting goal that we think we kind of have a unique place in the ecosystem that we can help empower others to-

    12. ET

      Yeah

    13. MZ

      ...make fast progress on that.

    14. ET

      So there, there, there's obviously like plenty of, uh, advantages to decentralization from a management communication overhead and so forth. And so, like, what are you trying to add by adding this kind of new layer/unification on top? Like, what, what are the outputs, and then I guess what are the complexities to that? 'Cause that's, um... I'm sorry to ask a CEO question. [laughs]

    15. MZ

      No, no, I, I mean, I'm, I'm like obsessed with this stuff.

    16. VA

      Ask for a friend.

    17. ET

      Yeah. I'm obsessed.

    18. VA

      We, we think about this-

    19. ET

      You, you wanna go for it, and then I can jump in.

    20. PC

      Yeah, so there are obviously amazing groups doing frontier AI and a lot of groups doing, uh, great frontier biology. And where we think we can do uniquely is actually tie these two together. And we

  15. 30:3233:36

    Why Biohub matters

    1. PC

      are, we've funded data sets, we've built data sets. We're like building the instrumentation now to be able to look at the cell, whether it's, you know, for at the tissue cell-cell communication, our cryo-EM, where we can look at the cell at nearly atomic level. So we have the ability to not only build the data sets, but actually shape and form them the way we want based on what we see as necessary to complement the existing body of knowledge. And so we have amazing teams doing that work, and we're building these AI models. And so what w- the reason to do it together is then we can actuallyComplete the flywheel. Like, you know, the m- model is looking like it has some gaps and blind spots in this area. Okay, who do we talk to? How do we build-

    2. ET

      Yeah

    3. PC

      ... um, the next data set? And, you know, we're seeing this in the lab, like the metadata is gonna be so rich that we can feed back into the way that we do this modeling.

    4. ET

      Yeah.

    5. PC

      And so if we can close that loop, which is our goal in bringing everyone together, um, it's, uh, I think it's gonna be incredibly powerful and it's, it's more than, it, it's more than just like, you know, writing down a spec and saying like, "Please deliver this." Like, these people need to be sort of working shoulder to shoulder and shaping, uh, each other's work for this to actually, um, be the, a more and more accurate model of how the human cell works.

    6. ET

      Well, you know, that's so interesting 'cause that is exact- Like, that's just been the biggest surprise in the industry for us in AI world, like forget biology for one second, is that the domain specific models have been like super interesting.

    7. MZ

      Mm-hmm.

    8. ET

      Like the, the, the original thesis was like there's just some AIs are gonna get so smart-

    9. MZ

      Mm-hmm

    10. ET

      ... they're gonna be smarter than everybody at everything. But, um, like on video models, like every video model is best at something but not everything. [chuckles]

    11. MZ

      Mm-hmm. Mm-hmm.

    12. ET

      And so knowing what problem you're solving actually turns out to be sort of ironically very important in AI, um, because you can actually get to a way better result-

    13. PC

      Yes

    14. ET

      ... if you put the two together. Like, yeah, we're, we're seeing that over and over, over again, uh, in a way that, that is, I would say-

    15. MZ

      Yeah

    16. ET

      ... very counterintuitive to the whole narrative kind of going into it.

    17. PC

      In biology, it used to be the, or at least, you know, one assumption was all the data sets aren't on the internet.

    18. ET

      Yeah.

    19. PC

      And so part of the reason you need a domain specific model is that the data sets are not public.

    20. ET

      Yeah.

    21. PC

      You guys are kind of bucking that trend too by creating a lot of open source-

    22. ET

      Mm-hmm

    23. PC

      ... access to the data and then even then it sounds like you're betting, you know, on the trend that we're seeing in other industries.

    24. ET

      Well-

    25. PC

      But still there will be nuance in how you annotate that data, curate that data-

    26. ET

      Well, and how you talk to a scientist, right? Like, so-

    27. PC

      ... and refine the model

    28. ET

      ... 'cause you have to not only know the-

    29. PC

      Yeah

    30. ET

      ... the data and the model and so forth, but like the conversation is what we keep-

  16. 33:3635:34

    The importance of interface design in democratizing scientific discovery

    1. ET

      exciting.

    2. PC

      And the user interface is actually really important. Um, y- you talked about, uh, you guys have a founder who's using Cell by Gene. That user interface was intentionally designed to not need to have a computational or really a very deep biological background to be able to use, because you want people coming from different fields-

    3. ET

      Right

    4. PC

      ... to look at the problem.

    5. ET

      Asking questions.

    6. PC

      It's like, "Look here. Help us solve problems here." And so building the user interface in a way where it's not a very high barrier to entry to be able to poke around and learn something, and bring knowledge back to your work, that's intentional, and we're really hoping when we build these virtual models, um, that, uh, we get to a place where we can allow a lower and lower barrier entry for people to say like, you know, like, "I have some knowledge about this. Maybe I can contribute." Um, a very pertinent example is turns out, I think immunology has a ton to do with neurodegeneration, right? But-

    7. ET

      Seems like immunology is behind all diseases-

    8. PC

      Everything

    9. ET

      ... so might be part of your century vision.

    10. PC

      Uh, so you need to be able to allow the immunologists to come in-

    11. ET

      Right

    12. PC

      ... and understand neurodegeneration and understand how their world fits in.

    13. ET

      Right.

    14. PC

      And so the more you lower the barrier to entry allows people to actually think in a sort of truly collaborative and interdisciplinary way.

    15. ET

      So will the Biohub grow as a team? Like, will you employ more people at the Biohub proper, or are you moving towards more of a network model with more sites, more labs, more community driven data sets? Like which, which is the thrust, or maybe it's both.

    16. MZ

      Probably a little of both, and we've added new Biohubs over time. Um, and then we're also building up more of this like central AI team.

    17. ET

      Cool.

    18. MZ

      So, um, but I don't-- I think that these organizational questions of how do you set this up are fascinating.

    19. ET

      [laughs]

    20. MZ

      And a lot of our approach is sort of informed by what the rest of the field is doing, because

  17. 35:3440:38

    How Biohub encourages cross-functional collaboration

    1. MZ

      I-

    2. ET

      Mm.

    3. MZ

      You kind of think about science as it's this portfolio, right? Society has a portfolio of stuff that it's trying to do.

    4. ET

      Mm.

    5. MZ

      And as in terms of philanthropy, you wanna be the most additive that you can be by trying to figure out what else is underrepresented. So science by default is very decentralized, right? It's like kind of the-

    6. ET

      Super, yeah

    7. MZ

      ... the way that granting has worked, the way that I think scientists by default want to work.

    8. ET

      Mm-hmm.

    9. MZ

      Um, so I think a lot of what we've found is that figuring out ways to encourage collaboration in, um, ways that otherwise seem very simple, but weren't happening before, can unlock a lot of value.

    10. ET

      Right.

    11. MZ

      So the very first Biohub, what we did, there were two kind of interesting things. One was, it was this collaboration between UCSF, Stanford, and Berkeley, and there are all these really smart people at all these different places who previously, I guess in theory, they could have figured out a way to work together, but there was not really a formal construct for them to do that.

    12. ET

      Yeah.

    13. MZ

      And this just allowed a lot more collaboration.

    14. PC

      Mm-hmm.

    15. MZ

      The other one is cross-discipline, basically having biologists sit next to engineers, and this view that like these two disciplines are-

    16. ET

      Yeah

    17. MZ

      ... things that need to, um... And I, I don't know. I mean, I'm, I'm sure, you know, you've seen this in a lot of, in, in a lot of the companies, but like it's, uh, there's so many interesting-

    18. ET

      [chuckles] And the companies, they always set them apart.

    19. MZ

      Well, it's interesting. [laughing] No, it's interesting how many organizational questions or problems you can fix just by having two teams sit together.

    20. ET

      Oh, yeah.

    21. MZ

      Right? It's like it doesn't matter what the org chart is or like whatever.

    22. ET

      Yeah.

    23. MZ

      It's like you guys need to sit next to each other and-

    24. ET

      Yeah

    25. MZ

      ... until you get this thing to work. And, um-That's something I really believe in, so-

    26. PC

      Then you have time.

    27. ET

      Yeah.

    28. PC

      You have 10 to 15 years.

    29. ET

      Well, no, it's all like communication-

    30. PC

      Figure out how it works, yeah

  18. 40:3844:43

    Looking ahead: the broader impact of AI on biotech

    1. ET

      about for the future or maybe even principles or a North Star that's gonna guide how you guys g-grow and evolve g-going forward?

    2. PC

      You know, it's been really interesting in the past ten years because I actually spent the first few years completely envious of people working for for-profit companies.

    3. ET

      Mm-hmm.

    4. PC

      Because there's so much clarity. Like the market will tell you whether or not it's private or public, will tell you if you're doing a good job.

    5. MZ

      If they think you're doing a good job.

    6. PC

      If they think you're doing [laughs] a good job.

    7. ET

      [laughs] They're not always right.

    8. MZ

      Yeah. [laughs]

    9. PC

      They're not always right.

    10. ET

      I know, it's a bit different. [laughs]

    11. MZ

      [laughs]

    12. PC

      But I was still envious 'cause that was-- I was like, I craved that feedback like, "Am I doing a good job?" And, you know, ten years in, you know, the reason why we're doubling down on biology is like, not only did we achieve what we said we were going to do a-and when we set out to set out on these projects, it actually delivered more than we thought we were going to. And I was like, "Okay, that's a signal I can latch onto," and like that's a signal I-- we can really continue doubling down and doing more of that. And so I think it's, uh, continuing to tolerate the early ambiguity when you're like, "Okay, I'm gonna do more of this." Um, and, uh, and being patient, but, uh, uh, being willing to have a long time horizon, but be impatient at the same time.

    13. VA

      Mm-hmm.

    14. PC

      'Cause it's all those iterations along the way that have sort of allowed us to get to this place where, you know, to get lucky, ready, having built data, data sets to take advantage of AI and large language models, that's because of all the work that we have been doing. And so being able to continue moving forward in this ambiguity and sometimes lack of signal on a big goal, like I think we sort of set the DNA for that.

    15. VA

      Amazing.

    16. PC

      Oh, no pun intended.

    17. ET

      [laughs]

    18. MZ

      Yeah. But we get to see how many people use the tools-

    19. PC

      Blueprint

    20. MZ

      ... and, and the feedback. Yeah. Yeah. Yeah.

    21. ET

      Yeah. You have customers, which is pretty cool.

    22. MZ

      Yeah.

    23. PC

      Yeah.

    24. ET

      For philanthropy, like that's awesome.

    25. MZ

      Yeah. No, it's, it's one of the-

    26. ET

      Yeah

    27. MZ

      ... fun things about building tools-

    28. ET

      Yeah

    29. MZ

      ... is like you kinda get to see-

    30. ET

      Yeah

Episode duration: 44:43

Install uListen for AI-powered chat & search across the full episode — Get Full Transcript

Transcript of episode YnV8pgHtO5Y

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.

Add to Chrome