Andrew Saxe Launches New AI Safety Research Lab 'Principia' in London
A top neuroscientist is building a team to crack open AI's black box.
Deep Dive
Andrew Saxe, a prominent computational neuroscientist, is launching Principia, a new research lab in London focused on developing foundational theory for modern AI systems. The lab aims to create mathematically clean, interpretable models that explain learning dynamics and generalization in neural networks, with direct relevance to AI safety. They are hiring 2-3 research fellows for 6-month contracts to work on this agenda, starting immediately.
Why It Matters
This represents a major push to build predictive theory for AI behavior, which is critical for managing risks and ensuring alignment.