After a decade building Andreessen Horowitz’s life sciences practice into a nearly $4 billion operation, Vijay Pande walked away last June. The Stanford professor turned investor, best known for creating the Folding@home distributed computing project, co-founded VZVC with longtime investor Zach Werner. The firm is deliberately small: a handful of concentrated bets per year, no associates, and AI woven through its daily operations.
Pande’s pivot reflects a specific thesis about where biology is headed. He argues the field is shifting from a science of discovery to something that can be engineered, with AI now capable of identifying drug targets, designing molecules, and eventually transforming clinical trials. That last part remains stubbornly expensive. Pande points out a drug’s probability of success from first trial through third is just 20%, with each trial potentially costing hundreds of millions of dollars. The failure isn’t typically bad biology — it’s that animal models like mice are poor predictors of human responses. AI models, even imperfect ones, would clear that bar easily.
The bigger prize, he says, is precision medicine. Today’s doctors guess: they prescribe a drug, wait, then try another. Pande wants the first drug to be the right one, comparing a patient’s blood values against their own baseline rather than population averages. Genomics alone won’t get there — your genome is the blueprint for a house on day one, not the house as it stands now. Proteomics and robotic measurement automation matter more for understanding disease in real time.
One unresolved tension hangs over all of this: biological data can’t be scraped off the internet like text. Every company builds its own walled-off dataset, which raises uncomfortable questions about who actually benefits from AI-driven medicine. That fragmentation feels like the inverse of what made Folding@home powerful — millions of machines contributing to a shared scientific resource. If AI in biotech stays siloed, the promised advances may arrive slower and less equitably than the hype suggests.
