Before OpenAI’s GPT‑3 ushered in the era of foundation models, companies built task‑specific natural language models from scratch, training each on massive, bespoke datasets. Today most organizations start with a general‑purpose model—GPT‑4, Claude, or Llama—and fine‑tune or prompt it for their needs. General Intuition’s CEO Pim de Witte believes robotics will follow the same trajectory.

From Specialized Datasets to General AI

De Witte argues that instead of amassing huge real‑world robot datasets, the industry should concentrate on high‑quality simulated data that can produce a foundation model capable of transferring intuition about movement and interaction across environments. “Many companies are still doing a lot of specialized work focused on individual embodiments, environments, and robots,” he told TechCrunch. “That work will soon become redundant.”

Leveraging Video‑Game Action Data

General Intuition trained its model on millions of hours of video‑game footage, capturing details such as which controller buttons a human pressed and when. Lead investor Vinod Khosla and de Witte both stress that this “action data” is the key to teaching a robot human‑like spatial‑temporal reasoning.

Funding Milestone and Live Demonstrations

Last month the startup raised $320 million, valuing it at $2.3 billion. In a live demo, the model not only played a video game for hours but, after fine‑tuning on just eight minutes of real‑world robot data, powered a quadrupedal robot. “The robot was able to zero‑shot using only a front‑facing camera, with no other sensors, in an office where dynamic objects and people were introduced,” de Witte said, calling the result a “big surprise.”

Implications for the Robotics Ecosystem

General Intuition’s endgame is not to manufacture robots themselves but to become the foundational model for physical AI— a base that other robotics firms can build upon. As de Witte puts it, “We’re not going to build a self‑driving car company. We’re going to make it ten times easier for the next person to build a self‑driving car company.” This paradigm could dramatically accelerate innovation, lower entry barriers, and reshape investment dynamics across the sector.