General Intuition Raises $320M to Train AI Agents Using Video Game Data
General Intuition has raised $320 million to advance its approach to training AI agents using data derived from video games. The company believes that action data captured from gameplay can help AI systems develop something closer to human intuition, potentially bridging the gap between virtual training environments and real-world applications.
The startup's strategy centers on scaling AI trained on vast quantities of gameplay footage, using the diverse range of scenarios and decisions present in video games as a training ground for more adaptable AI agents. By exposing AI systems to millions of hours of in-game action, General Intuition aims to develop agents that can reason and respond more naturally to novel situations.
The $320 million funding round will support the company's efforts to expand its data infrastructure and model capabilities. General Intuition's approach represents a growing trend in AI development where researchers seek to move beyond traditional training methods that rely heavily on curated datasets, instead exploring how interactive and experiential data can produce more flexible AI behavior.
The bet on video game data reflects broader industry interest in using simulated or virtual environments to train AI systems for real-world tasks, from robotics to autonomous decision-making applications.