General Intuition raises $220M at $6.2B valuation for world models
General Intuition raised $220 million at a $6.2 billion valuation to expand synthetic video world models and test a commercial version.
According to SiliconANGLE, the investment comes a year after General Intuition spun out of Medal B.V., a video-sharing startup. Medal provides a free tool that consumers use to capture and record video game footage, and General Intuition uses that footage to train its AI models.
Developers train AI models by providing examples of tasks they will perform. For instance, an algorithm designed to power a robotic arm might be given footage of robotic arms welding car parts. Manually recording such footage can be time-consuming, and General Intuition aims to ease that process.
The company released its newest algorithm, MIRA, in June. It says MIRA is more capable than earlier models in several respects. Traditional video generators can render only clips of a few seconds or minutes, which prevents them from generating synthetic videos of long-running factory automation workflows and limits their usefulness for robot developers. MIRA, by contrast, can "run infinitely without diverging," according to the company. It can also render complex scenes in which multiple fast-moving objects interact, an ability useful for tasks such as training robots to avoid collisions.
MIRA generates footage efficiently. General Intuition says it can render 20 frames per second at a resolution of 720 by 576 pixels using a single B200 graphics card. The model's efficiency partly comes from latent diffusion, which runs calculations on a latent space, a compressed representation of video frames that requires less memory. Shrinking a file's memory footprint reduces processing time. MIRA's size also factors in: it comprises only 5.6 billion parameters, a fraction of the number in frontier models.
MIRA is a research demonstration rather than a commercial training data generator. In its current iteration, the model can generate synthetic footage of only a single video game. General Intuition says MIRA's efficiency and rendering quality make it a "stepping stone to physical AI."
The company is testing a commercial version of its technology with a limited number of customers working on robotics, simulation and entertainment use cases. It opened a waitlist for the offering alongside the funding announcement. The company said it will use the capital to hire more AI researchers.