Anthropic says Claude 'leads' 26 percent of its AI R&D work
Anthropic reports that its own AI model Claude is autonomously leading a significant share of its AI research and development, raising concerns about recursive self-improvement and loss of human oversight.
Anthropic reports that its own AI model Claude is autonomously leading a significant share of its AI research and development, raising concerns about recursive self-improvement and loss of human oversight.
Auto-scored: A frontier lab disclosing that its AI leads over a quarter of its own R&D signals a real capability jump and monitorability regression.