Jeff Dean's Discovery Loop seeks $1B at $10B valuation
The startup from Google's most famous engineer, announced barely a week ago, now carries a price tag that puts it in the same league as the model labs — and the money is betting on a different game entirely.
Jeff Dean's Discovery Loop is in talks to raise $1 billion at a roughly $10 billion valuation, people familiar with the matter told Business Insider — one of the largest early-stage rounds ever for a company that publicly introduced itself barely a week ago. The departing Google chief scientist, who spent 27 years at the company, has not commented, and the people familiar cautioned that details of the round could still change. What is already public, from a Discovery Loop press release on August 5, is the investor lineup: the initial round is led by Radical Ventures and Khosla Ventures with participation from Lightspeed, Kleiner Perkins, and Doerr Capital.
The thesis behind the valuation is unusual even by frontier-lab standards. Discovery Loop is structured as a public benefit corporation whose mission, as Dean described it on X, is "to automate machine learning, science, and engineering to accelerate discoveries and progress" — attacking hard problems through the "parallel execution of thousands of experiments" rather than through a single bigger model. That is the same pitch from the founding announcement we covered last week — Google exodus: Jeff Dean founds Discovery Loop — with co-founders Sanjay Ghemawat, Oriol Vinyals, and Quoc Le, a founding team whose fingerprints are on Search, Ads, Gemini, and Gmail. Google keeps a stake and signed on as the founding cloud partner.
The number puts Discovery Loop in a striking peer group. Former DeepMind researcher David Silver raised $1.1 billion at a $5.1 billion valuation for Ineffable Intelligence this year, and Reflection AI reportedly raised at a $25 billion pre-money — the market is now routinely pricing fresh AI science startups above most public software companies. What makes the $10 billion figure notable is that it prices an unproven process — automated experimentation — as if it were a shipping product, which says more about investor hunger for whatever comes after the chat era than about any public benchmark. If Dean's loop works, it is not another model lab; it is infrastructure for discovery itself, and at this valuation the market is already treating it that way.
What to watch: whether the round closes anywhere near $10 billion, and whether Google's stake and cloud partnership survive the transition from co-founder to competitor.
Would you back a $10 billion bet that AI can automate the scientific method? Tell us in the comments.
