Bloomberg reported Friday that Stanford professor James Zou is raising about $100 million at a roughly $1 billion valuation for a startup called Human Intelligence, building AI models for research on the human body. [1] The Next Web's Saturday wrap added the credentials: Zou's lab produced EchoNet, an FDA-cleared cardiac AI, plus a Nature-published Virtual Lab that designed novel nanobodies and a Virtual Biotech multi-agent framework that annotated 56,000 clinical trials. The Next Web also flagged that "the company's name, valuation, and raise amount have not been independently confirmed" outside Bloomberg's sourcing. [2]

That caveat is the story. AI drug discovery received $11 billion in Q1 2026 alone. [2] The price tag on Human Intelligence is not what one paper or one model is worth — it is what an academic with a track record of clinically deployable systems is worth before the foundation model exists. A "physiology foundation model" is, in practice, a wager that the same transformer architecture that ate language can eat the human body's signals.

The discipline check matters. No AI-discovered drug has completed a pivotal Phase III trial. EchoNet is real; Virtual Lab is published; the human-physiology model is unbuilt. Northwestern's printed-neurons paper this month is in the same news cycle and at the opposite end of the maturity curve. [3] Sunday's read: venture capital is paying unicorn prices for principal-investigator pedigree. The biology bet is now a labor-market signal.

-- KENJI NAKAMURA, Tokyo