A startup tripling its valuation between funding rounds inside a single quarter is the story here, and the underlying asset behind that math is data: hundreds of millions of hours of gameplay footage paired with ‘action labels’ — records of exactly what buttons a player pressed and when. General Intuition is betting that this unglamorous dataset, mined from a video-clip-sharing app rather than scraped from the web, can teach AI agents to reason about space, time and embodiment in ways text corpora can’t.
General Intuition’s jump from a $2.3 billion valuation to a reported $6 billion pre-money mark in a matter of weeks says less about revenue and more about how scarce investors think usable physical-world training data has become. The round is reportedly oversubscribed, with Khosla Ventures and General Catalyst returning alongside new entrants Valor, Point72 Ventures and Seven Seven Six — and TechCrunch notes this would be Valor’s first AI lab bet since SpaceX, a pattern-match investors will read as a bullish signal on physical AI infrastructure, not just software.
Worth flagging: TechCrunch’s reporting is attributed to sources familiar with the matter, the deal is still being finalized, and the outlet says it has reached out to Valor to confirm — this is a developing story, not a closed transaction, and the $6 billion figure hasn’t been independently verified elsewhere. If it closes near that number, expect rival labs building on gaming, simulation, or robotics telemetry data to start pointing to their own ‘action label’ equivalents when they pitch their next round.
New York-based General Intuition, the startup building a foundation model that trains generalized AI agents to move through space and time, is in talks to raise funding at a $6 billion pre-money valuation from new investors including Valor Equity Partners, Point72 Ventures, and Seven Seven Six, according to sources familiar with the matter.
