AI 研发自动化
Using AI systems to perform research, engineering, and experimentation tasks, thereby accelerating progress on AI capabilities t
AI safety research organization METR published a report after an 11-week investigation inside Anthropic, finding the company is using AI to build more powerful AI, with model capabilities improving roughly twice as fast as a year ago and little sign of slowing. The findings have alarmed researchers about the risks of automated AI R&D, with METR's leadership urging society to brace for very rapid change.
Anthropic's S1 filing says it may 'imminently meet' its Responsible Scaling Policy threshold for dramatically accelerated progress from AI, defined as overall capability progress at twice the fastest pre-AI sustained rate and plausibly attributable to AI automating AI R&D. A public METR report and outside commentary treat the claim as substantive, though the underlying evidence and unredacted data remain undisclosed.