General Intuition, a New York startup building foundation models to teach general AI agents how to navigate through space and time, is in talks to raise funding at a pre-money valuation of 6 billion USD. The roster of new investors includes Valor Equity Partners, Point72 Ventures, and Seven Seven Six, with existing backers Khosla Ventures and General Catalyst also participating.
The startling figure lies in the speed: just in late June, the company raised 320 million USD at a 2.3 billion USD valuation. A few weeks later, the valuation has nearly tripled. Sources close to the deal say the round is currently oversubscribed.
What is drawing investor capital is the training data. General Intuition was spun out last October from Medal, a gaming clip-sharing platform led by CEO Pim de Witte: hundreds of millions of hours of gameplay paired with action labels, capturing which buttons a player pressed and when. Investor Vinod Khosla believes these action labels will be the key to the “emergence of intuition”—the model’s ability to generalize to tasks it was never trained on.
The new capital will be used to enhance the general model with a focus on robotics, along with compute infrastructure costs (in partnership with CoreWeave) and hiring. If the deal closes, it will be the first AI lab that Valor Equity Partners, a venture firm well-known for its investment in SpaceX, has backed since SpaceX.
Transparency note: The deal is still under negotiation; the information comes from unnamed sources and press reports (previously reported by the WSJ), with no official confirmation yet from the parties involved.
The bigger picture: while crypto capital this week focused on ETFs, AI venture funding is pouring into physical AI and world models—systems that interact with the physical world rather than just text. A valuation tripling in a matter of weeks is both a sign of intense heat and an indicator that valuations have run ahead of product evidence. Both can be true at once, but the gap between them is usually paid for in expectations.