Beyond the Screen: General Intuition’s $6B Bet on Physical AI Foundation Models
General Intuition is developing foundation models that teach AI agents to navigate complex 4D environments, bridging the gap between digital intelligence and physical movement.
The quest for Artificial General Intelligence (AGI) has long been dominated by Large Language Models (LLMs) that process text and code. However, a new frontier is emerging: Physical AI. General Intuition, a startup that recently secured a $6 billion valuation, is leading this charge by building foundation models specifically designed to train AI agents on how to move through space and time. Unlike traditional AI that lives in a vacuum of data, Physical AI requires an understanding of physics, spatial relationships, and temporal dynamics.
Investors like Valor and Point72 are betting that the next great leap in technology won't happen on a screen, but in the physical world. General Intuition’s approach focuses on 'generalized AI agents'—software entities capable of operating a variety of mechanical forms, from robotic arms in factories to humanoid assistants in homes. By mastering the nuances of physical interaction, these models aim to solve the 'Robotics Transformer' challenge: creating a single brain that can control many different bodies.
This shift represents a fundamental change in how we develop autonomous systems. Instead of hard-coding specific behaviors for specific tasks, Physical AI enables machines to learn through observation and simulation, much like how a human learns to navigate a room. As capital pours into this sector, the boundary between software and hardware continues to dissolve, promising a future where AI is truly embodied.
Source: TechCrunch
