Apr 2025
Amodei argues for accelerating work to understand how advanced AI systems operate internally.
Connects mechanistic-interpretability research to his proposed approach to AI safety.
Explore on Humanity Unfolds
