This startup thinks robotics is about to have its ChatGPT moment
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- General Intuition raised $320 million at a $2.3 billion valuation last month, with Vinod Khosla as lead investor, on the thesis that physical AI will follow the LLM shift from specialized to general-purpose foundation models.
- CEO Pim de Witte argues that "the generalization of the model itself is the product" and that companies collecting "hundreds of thousands or millions of hours of real-world data" will find that work redundant once general models emerge.
- General Intuition trained its foundation model on millions of hours of video game data — including which controller buttons humans pressed and when — to teach human-like spatial-temporal intuition rather than collecting real-world robotics footage.
- The company demonstrated its current model can zero-shot power a quadrupedal robot using just a front camera and eight minutes of real-world fine-tuning data, a result de Witte called "a very big surprise."
- General Intuition is not building its own robots but positioning itself as the base model layer for other companies, with de Witte saying the goal is to "make it 10 times easier for the next person to build a self-driving car company."
Why it matters: General Intuition's bet attacks the data-collection moat that incumbent robotics and self-driving companies built on years of expensive real-world footage. If a game-trained model can zero-shot a quadruped with just eight minutes of real-world data, the cost floor for launching a physical-AI startup collapses — and the $2.3B valuation signals VCs are pricing robotics as a platform layer, not a collection of bespoke robot builders.
