Anthropic says it may soon meet its threshold for AI-accelerated research
Anthropic has not designated any current model as exceeding that threshold and is considering human review of some research decisions.
TLDR
Anthropic's estimates and expectations for future models suggest it may soon meet its Responsible Scaling Policy's definition of “dramatically accelerated progress attributable to AI.” The company is considering requiring human approval for certain research decisions as models perform more of that work. It does not expect to safely use Claude to fully autonomously design and build successive generations of frontier models without meaningful human supervision or understanding, if that becomes technically possible.
Combined views
28K
14 Sources, first seen ago



