Anthropic warns AI could outpace human control

Anthropic co-founders Jack Clark and Marina Favaro warn AI may achieve recursive self-improvement, allowing systems to design successors and increasing risks of human loss of control. The company notes slowing AI development could give time to address implications. David Sacks criticizes Anthropic for advocating government oversight of AI labs. Mitt Romney calls AI safeguards a national priority citing risks of unemployment, surveillance, and extinction. Stanford professor Andrew B.