scripod.com

Priscilla Chan and Mark Zuckerberg: Frontier AI + Virtual Biology To Solve All Diseases

In this conversation, Priscilla Chan and Mark Zuckerberg reflect on a decade of scientific innovation through the Chan Zuckerberg Initiative, now advancing one of the most ambitious frontiers in modern science: merging artificial intelligence with biology to reimagine how we understand and treat disease.
The discussion traces the evolution of their scientific mission from broad philanthropy to a focused push on AI-driven biomedical discovery. Central to this vision is the Biohub, where interdisciplinary teams are building comprehensive models of cells and the immune system by integrating vast biological datasets with cutting-edge AI. A major bottleneck remains data acquisition, prompting investments in high-resolution imaging and custom-built microscopes to digitize biological structures at scale. AI models like rbio1 and VariantFormer are being trained on this data to simulate cellular behavior, test hypotheses, and accelerate research that was once limited by wet lab experimentation. The team emphasizes open science, collaboration, and computational modeling as key to de-risking bold research. With the integration of Evolutionary Scale’s protein modeling expertise, they aim to scale discovery and develop a virtual immune system capable of engineering immune cells for precise disease intervention. Ultimately, the goal is to shift medicine from reactive to proactive—enabling early detection, personalized treatments, and a future where curing all diseases is grounded in data, models, and cross-disciplinary innovation.
09:28
09:28
AI is the next step in software 'eating the world', now transforming biology.
10:45
10:45
Hierarchical modeling starts with proteins, then cells, leading to a virtual immune system.
13:58
13:58
The billion-cell project is much faster and cheaper than previous efforts
21:20
21:20
A recent diffusion model can generate synthetic cells for biological research.
26:33
26:33
AI models will first help generate hypotheses before replacing wet labs
28:47
28:47
Alex Reeves will lead the Evolutionary Scale program, combining AI and biology to advance frontier protein models.
30:15
30:15
AI models could predict how genetic variants affect cells and disease pathways.
45:21
45:21
CAR T cells exemplify how engineered immune cells can target cancer, paving the way for broader applications in disease prevention.
51:44
51:44
Curing diseases requires collaborative AI-biology networks