Podcast Summary: Latent Space — Priscilla Chan and Mark Zuckerberg: Frontier AI + Virtual Biology to Solve All Diseases

Date: November 6, 2025
Guests: Priscilla Chan (CZI), Mark Zuckerberg (CZI & Meta)
Hosts: Alessio (Kernel Labs), Swyx (Latent Space)

Episode Overview

This episode marks the 10-year anniversary of the Chan Zuckerberg Initiative (CZI) and dives deep into the intersection of AI and biology. Hosts Alessio and Swyx explore CZI’s mission to "cure, prevent, or manage all diseases" with Priscilla Chan and Mark Zuckerberg, focusing on foundational and frontier biological research augmented by next-gen AI models and in-house scientific tool development. The conversation traverses the history, impact, technical approaches, challenges, and future vision for virtual biology and disease prevention.

Key Discussion Points & Insights

1. Origins and Philosophy of CZI

• CZI vs. Gates Foundation: CZI started with an experimental approach, learning by doing, rather than focusing immediately on translational medicine as the Gates Foundation does ([00:51]).

  "We should just kind of dig in and start doing a few different iterations on it and see what we enjoy and where we think we can have an impact..."
  — Mark Zuckerberg [00:51]

• Unique Role: CZI emphasizes building research tools and fostering fundamental science, rather than conventional grant-giving ([04:26], [10:56]).

  "A lot of what we're doing is actually building up these institutes and building labs to do that kind of research ourselves..."
  — Mark Zuckerberg [04:26]

2. Mission to Cure, Prevent, or Manage All Diseases

• Ambition and Skepticism: Internally, the mission is seen as realistic; AI experts see it as inevitable, while biologists approach it more cautiously ([07:18], [09:00]).

  "When you ask AI people, it’s like, that should be really easy. Why are you so unambitious that you’re shooting for just the end of the century?"
  — Mark Zuckerberg [07:35]

• Ecosystem Approach: CZI’s Biohub promotes collaboration among scientists, AI researchers, engineers, and physicians, breaking traditional silos ([10:56], [14:15]).

  "It is... amazing how much progress you can make if you just have people from different disciplines sit together."
  — Mark Zuckerberg [10:56]

3. Data Creation, Tooling, and Modeling

• Cell Atlas and Data Growth: The Human Cell Atlas project seeded large-scale, collaborative data collection; recent projects collect exponentially more data in less time
  ([14:15]).

  "We now have one of the largest corpus of RNA transcriptomes. 125 million cells cost a lot of money. And the really cool thing... if we could seed the effort and make it easy for people to contribute, it happened."
  — Priscilla Chan [14:15]

• Custom Instruments & Imaging: Most cutting-edge microscopes are custom built. Imaging speed, dimensionality (spatial, time), and data fusion remain bottlenecks ([15:46], [16:13], [17:45]).

  "...with the cryo model it will get fast again and you just have to repeat it."
  — Priscilla Chan [14:15]

4. Virtual Biology and Iterative Model Development

• The Virtual Cell & Immune System: Stepwise modeling—protein → cell → system—mirrors advances in AI, with current work just at the beginning ([10:56], [45:12]).

  "You kind of need systems that understand data at all different levels... and then... you build a richer and richer model of how these cells work."
  — Mark Zuckerberg [10:56]

• AI Validity Loops: Unlike rapid LLM testing, AI in bio requires wet lab cycles for feedback, which are still orders of magnitude slower ([25:08], [26:06]).

  "You have to actually take it to the wet lab, run the experiment, find out if it actually happened as predicted, and feed it back into the model."
  — Priscilla Chan [25:08]

5. Biohub Evolution and Announcements

• Frontier Lab Model: Analogous to AI "foundation models," the Biohub unifies top talent (notably, the Evolutionary Scale team led by Alex Reeves, joining CZI) for deep AI-biology convergence ([28:42]).

  "I think it’s sort of an interesting decision... to have the AI person basically be running the overall program partnering with these leading biologists..."
  — Mark Zuckerberg [28:50]

• Compute as an Enabler: CZI built large compute clusters specifically for biological research—an atypical move in the science world ([29:48]).

6. A Vision for Precision and Proactive Medicine

• Clinical Translation: The real milestone is clinical impact. One major anticipated advance: identifying the effect of genetic variants with precision ([30:36]).

  "That is actually the future of medicine, where we think about each one of your biology based on your genetics, your exposure, and how it predisposes you or not to disease."
  — Priscilla Chan [30:36]

• Doctors’ Evolving Role: As models advance, physicians will focus more on care, interpretability, and guiding patients, rather than routine diagnosis ([36:59], [37:14]).

  "Care and compassion and sort of walking patients through understanding, I think understanding why leads to trust in both the science and in the clinical pathway."
  — Priscilla Chan [37:14]

• Proactive vs. Reactive Healthcare ([38:29]):

  "Everyone wants the health system to be more proactive and less reactive.... The goal... is to be much more proactive about this."
  — Mark Zuckerberg [38:32]

7. Philosophy of Scale and Timeframes

• Is 'Ending Death' the Goal? Focus is on maximizing quality and length of life, not necessarily radical life extension or 'ending death' ([39:53]).

  "I'm a pediatrician. I think about babies and like very sad things happen to very small people. And like, I think a lot about that and how do we like maximize life quality..."
  — Priscilla Chan [40:23]

• Acceleration Depends on AI: Ultimate pace is closely tied to AI progress; biology must keep producing frontier datasets and tools ([48:37], [49:47]).

  "If we're predicting... whether it's going to take 10 or 20 or 40 years, that is probably more a function of the pace of AI development than it is a pace of the pure biology side."
  — Mark Zuckerberg [48:37]

• Collaborative Science Culture: Large-scale, unglamorous data work (e.g., the 120 millionth cell) is essential, necessitating new models of scientific reward and collaboration ([49:47]).

  "We need to be continuing to push the research and the methodologies. And I want to say that the cell atlas was not glamorous work...."
  — Priscilla Chan [49:47]

Notable Quotes & Memorable Moments

• "Believing is the first step." — Priscilla Chan [10:15]
• "I think you want the models... to basically build up different levels of abstraction and pattern matching. And that's here too." — Mark Zuckerberg [44:16]
• "We need lots of people coming together to do this work." — Priscilla Chan [53:04]
• "Check out the models. They're early, but I think it's kind of an interesting sense of where things are going, and we'd love feedback on it..." — Mark Zuckerberg [52:47]

Important Timestamps

• 00:51 – CZI’s iterative, learn-by-doing philosophy in philanthropy
• 04:26 – Tool building in science; differences from traditional grant-giving
• 10:56 – Breaking down academic silos and enabling mixed-discipline teams
• 14:15 – Success of the cell atlas and community data efforts
• 25:08 – The necessity and slowness of wet lab validation cycles for AI models in biology
• 28:50 – Bringing in the Evolutionary Scale team; AI-first leadership in biology
• 30:36 – The vision of precision medicine: deeply personalized disease modeling
• 36:59 – The evolving, increasingly patient-centered and empathetic physician
• 45:12 – Virtual immune system: opportunities, clinical implications, and technology
• 48:37 – Timeline questions; pace of AI as primary determinant
• 52:47 – Call to action: try the models; collaborative plea to the global research community

Conclusion & Call to Action

CZI is pushing the boundaries of what’s possible in fundamental bioscience by uniting AI and biology experts, building bespoke tools and datasets, and embracing collaborative, long-term frameworks that break from academic tradition. Clinical impact remains the north star. The guests urge biologists, engineers, and AI practitioners to explore CZI's models, participate in the ecosystem, and help build the tools and data that will underpin the next era of medicine.

"Let's do this together." — Priscilla Chan [53:04]

Explore models and learn more at: https://latent.space