General Intuition is a frontier research lab based in New York, founded in 2025. The company builds large action foundation models, training them on billions of action-labeled video clips sourced from tens of thousands of games. Its stated goal is to teach machines a "general intuition of reality" - the ability to perceive, predict, and act across both virtual and physical environments.
The lab spun out of Medal, described as the world's largest platform for gaming clips, where millions of gamers capture and share billions of gameplay moments each year. This pipeline provides the training data at scale for General Intuition's models, which are designed to learn from video how to anticipate and improvise, with the aim of transferring capabilities from virtual to physical domains.
General Intuition operates as a public-benefit corporation. Its technical work spans several domains, including action prediction, world modeling, embodied AI, perception of space and time, and machine learning from video data. The company has stated a commitment to collaborating with rather than competing against the gaming industry and creative professionals.
