Jun 10, 2026 · 56m · no-priors
“Curing All Disease by next century is too conservative" - Mark Zuckerberg
Mark Zuckerberg · 18m spoken Priscilla Chan · 13m spoken Alex Rives · 11m spoken Sarah Guo · 5m spoken Elad Gil · 4m spoken
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
Mark Zuckerberg, Priscilla Chan, and Alex Rives discuss the Chan Zuckerberg Biohub's mission to accelerate biomedical discovery by uniting frontier AI with high-throughput wet-lab biology. They outline their open-source philanthropic strategy, breakthroughs in protein modeling, and the roadmap toward building multi-scale virtual cell models to help cure, prevent, or manage all human diseases.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. The hosts hold 17% of the talking time here. How this is scored →
The hosts as informed peer 3.8 Guest teaching 3.3 Guest disagreement 0.9 The hosts pushing back 0.7
speaking balance: gold is the hosts, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 39:20 Mark Zuckerberg rejects centralized superintelligence
Zuckerberg directly dismisses the popular industry premise of a single centralized superintelligence solving science, arguing forcefully for decentralized tools in individual hands.
Hardest push from the hosts ▶ 16:58 Sarah Guo challenges the non-profit funding approachGuo questions why a philanthropic non-profit model was chosen over an ambitious venture-backed startup model given the guests' entrepreneurial backgrounds.
Biggest teaching moment ▶ 21:55 Priscilla Chan reframes disease prediction into foundational mechanismsChan politely corrects Gil's framing about picking disease targets by demonstrating that modern computational biology should model fundamental cellular systems rather than clinical categories.
The host holds their own ▶ 12:35 Elad Gil cites wet lab background to address biological stratificationGil brings his molecular biology PhD background into the dialogue to point out historical fractures between reductionist and systems biology approaches.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The hosts as informed peer | Guest teaching | Guest disagreement | The hosts pushing back | Why |
|---|---|---|---|---|---|---|
| Expanding the Biohub Model to Frontier Biology and AI | 0 | 2 | 1 | 0 | Mark Zuckerberg delivers an uninterrupted explanation detailing the Biohub model and how novel frontier biology is required to generate datasets for AI models. The hosts do not intervene during this monologue. | |
| From Single-Cell Atlases to AI-Driven Predictive Biology | 0 | 3 | 1 | 0 | Priscilla Chan explains the historical progression from single-cell transcriptomics and Cell by Gene to using LLMs to turn biology into an engineering discipline. Hosts listen quietly throughout. | |
| Alex Rives on Uniting Frontier AI and Biology | 8 | 3 | 1 | 2 | Host Elad Gil showcases his technical background with a PhD in biology, asking a detailed question contrasting cellular modeling with protein folding and explaining the divide between reductionist and systems biology. | |
| Mechanistic Interpretability and Unlocking Black-Box Protein Representations | 5 | 4 | 0 | 0 | Sarah Guo probes mechanistic interpretability in biological models, and Alex Rives elaborates on how protein language models develop emergent representations of structure and grammar. | |
| Why Open-Source Philanthropy Outperforms Startups for Broad Scientific Impact | 4 | 3 | 2 | 2 | Sarah Guo asks why open-source philanthropy is superior to venture-backed startups for this mission. Zuckerberg and Chan argue that open-sourcing democratizes tooling across rare diseases and removes short-term monetization constraints. | |
| Targeting Foundational Biological Systems and Personalized Interventions | 4 | 4 | 2 | 1 | When Elad Gil asks which specific disease areas will be impacted first, Priscilla Chan rejects the premise, explaining that their approach targets underlying biological systems and individual mechanisms rather than disease categories. | |
| ESM Fold Breakthrough and De Novo Antibody Generation | 3 | 5 | 0 | 0 | Alex Rives and Priscilla Chan break down the technical breakthroughs of ESM Fold, predicting 1.1 billion structures and designing de novo nanobodies validated via cryo-EM. | |
| Predicting Off-Target Toxicity and Empowering Rare Disease Cohorts | 7 | 4 | 1 | 2 | Elad Gil raises the practical bottleneck of clinical trial timelines, costs, and toxicity failures. Priscilla Chan and Mark Zuckerberg explain how transcriptomic models and patient registries alter trial recruitment and toxicology predictions. | |
| Mark Zuckerberg on Decentralization and Empowering Individual Researchers | 3 | 2 | 2 | 0 | Sarah Guo asks how open ecosystems compare between AI models and biological research. Mark Zuckerberg firmly pushes back against the concept of a centralized superintelligence, arguing for empowering distributed researchers. | |
| Recruiting Elite AI Talent to Mission-Driven Frontier Biology | 3 | 2 | 1 | 1 | Sarah Guo asks how Biohub competes for AI talent against commercial enterprises. Zuckerberg and Rives emphasize their unique integration of wet lab wetware with computing and a mission-driven focus. | |
| The Path to Virtual Cell Modeling and Constraint Management | 4 | 4 | 0 | 0 | Sarah Guo asks about the practical inputs, outputs, and constraints of building a virtual cell model. Alex Rives and Mark Zuckerberg outline the hierarchical modeling requirements and compute trade-offs. | |
| Exponential Progress, Five-Year Outlook, and Strategic Alignment | 5 | 3 | 0 | 0 | Elad Gil reflects on the exponential growth curve and closing the loop between digital predictions and wet lab experiments. Zuckerberg and Chan summarize their strategic pivot toward AI-led biological research. |
Statements from this episode (22)
Zuckerberg: Curing all disease by end of century is too conservative
“But still, I mean, people thought that by the end of the century was a stretch. Now I think it's, like too conservative.”
Zuckerberg: Science has become the primary focus of CZI
“Now it is basically the primary and main thing that we're doing, and we've expanded the original San Francisco Biohub to a handful now at this point. There's New York, there's Chicago.”
Zuckerberg: Biological AI requires wet labs because internet data is insufficient
“You want to build a frontier AI lab, but you need to couple that with a frontier biology effort that can do the work of basically being able to understand and get the data that you need to actually be able to build these models, because unlike language models,…”
Priscilla Chan: Human Cell Atlas is one of the largest single-cell transcriptome databases
“Us funding the Human Cell Atlas, which is now one of the largest databases of single-cell transcriptomes.”
Priscilla Chan: Cell by Gene powers many transcriptomic-based biological models
“And now Cell by Gene is a corpus of knowledge that a lot of the transcriptomic-based models are based off of and is used regularly by the scientific community.”
Rives: Next-Token AI Models Can Learn Biological World Models from Data
“We have these systems that can basically kind of predict the next token and they can, you know, learn, World models from that. They can learn biology from the data.”
Zuckerberg: Biological simulation requires hierarchical modeling from proteins to systems
“I mean, the, but you need to be able to understand the protein interactions in order to be able to understand how cells work. So you can't just go straight to cells in a way without understanding the protein modeling. And then if you're trying to understand so…”
Alex Rives: Protein language models emergently learn biological structure and function
“One of the classes of models that we train are these protein language models. So they're really, you know, it's trained on the codes of proteins, and so anything they learn about biology is, is kind of emergent. And we've seen that they can learn things like b…”
Alex Rives: Mechanistic interpretability will uncover biology inside protein models
“The hope is that you kind of really learn the underlying basis for how it's making the predictions, and so you open up the black box and you can actually understand kind of the biology that the model is representing.”
Zuckerberg: Building complex biology tools requires a 10-15 year non-profit horizon
“If you're building tools that are this complicated, you kind of want to have a 10 to 15 year time horizon on, on building out these efforts. And then the scale of capital required. I mean, I guess there's no rule that said that you couldn't do it as like an in…”
Zuckerberg: Biohub will not cure diseases directly but build tools for scientists
“And again, I mean, the theory isn't that we're going to cure the diseases. We're not. It's that we want to help accelerate the pace of progress for the whole scientific field.”
Chan: Curing disease requires uniting academia and industry via non-profit neutrality
“The sort of neutral nonprofit nature of our work actually helps harness more people to enter this effort. And to actually achieve the mission of, like, understanding the totality of human biology and to cure, prevent, manage all diseases, you actually do need …”
Priscilla Chan: Medicine currently lacks mechanistic understanding and relies on trial-and-error
“We really have no mechanistic understanding. We're saying, like, okay, you're kind of like these people that we studied, and this drug kind of impacts the pathway that we think is implicated. Let's try and see if anything happens. And time passes, and sometime…”
Rives: ESM Fold has predicted structures for over 1.1 billion proteins
“So we folded over 1.1 billion proteins and predicted their structures and identified kind of features connecting all of them through mechanistic interpretability.”
Rives: ESM Fold achieves SOTA in protein and antibody interaction benchmarks
“And it's really, you know, hitting state of the art across pretty much every structure prediction benchmark, and especially on protein, protein interactions and protein antibody interactions, which is really critical for therapeutic design.”
Chan: Single-Cell Atlases Enable Pre-Trial Prediction of Off-Target Drug Toxicity
“If you have a single cell atlas that looks at all the different cell types some of which actually were not predicted before we modeled them, you can start looking at which cells actually do have receptors for the target you thought you were exclusively targeti…”
Zuckerberg: Studying rare edge cases yields more insights than common diseases
“I think one of the interesting things from, you know, science and engineering is that often, you know, you can hit your head against the wall on the common problems, and in this case, diseases, but a lot of times you, like, learn a lot more about a system from…”
Chan: Self-Organized Patient Cohorts Advanced Gene Therapy in Under Five Years
“There's gene therapy that one disease group has moved forward over the course of, like, I want to say, like, three to five years rather than decades, and the speed is so fast. Because the patients themselves have organized the resources that a scientist or a c…”
Zuckerberg: A Centralized Superintelligence Will Not Solve Science
“Our vision is not that there's going to be like some central super intelligence that solves all of science.”
Zuckerberg: AI progress requires only a couple dozen researchers, not hundreds
“In order to make progress in AI, you don't need like many, many hundreds of AI researchers or thousands or anything like that. I think you can really make progress with you know, a very strong group of a dozen or a couple dozen people.”
Rives: Users are pairing ESM models with AI agents for automated design
“I think one of the really interesting things that we've been seeing is people kind of connecting it with agentic systems to just kind of do automated design and kind of just automate that, that whole process.”
Zuckerberg: Biohub leadership shifted from biologists to AI researchers
“Prior to Alex leading the effort, the previous leaders of the Biohub were basically primarily biologists who were interested in technology, right? And now I think we, this is the point where we really flipped that, right? Where, I mean, obviously you have a ba…”