Google DeepMind partners with CFS to control fusion plasma with AI
The collaboration aims to use reinforcement learning to stabilize tokamak reactors by 2027.
AI as an enabling technology for clean energy
Google DeepMind announced a strategic partnership on Tuesday with Commonwealth Fusion Systems (CFS), an MIT spinoff working to commercialize nuclear fusion energy. The collaboration focuses on deploying advanced artificial intelligence to control the superheated plasma inside a tokamak reactor. Maintaining the stability of plasma—which reaches temperatures hotter than the sun—requires millions of micro-adjustments per second to magnetic fields, a task that exceeds the capabilities of traditional algorithmic control systems.
DeepMind will utilize reinforcement learning models to predict plasma instabilities milliseconds before they occur. The AI learns to adjust the reactor’s magnetic …
Archive Access
This article is older than 24 hours. Create a free account to access our 7-day archive.