Anthropic warned in a revised regulatory filing that it may soon develop artificial intelligence capable of recursive self-improvement. The company pledged to maintain human supervision over future models to prevent automated development from outrunning safety measures.
FireBrief
FireBrief