Anthropic may have already crossed its own threshold for AI-driven acceleration
AI is dramatically accelerating AI research
A METR deep-access investigation of Anthropic found that AI capabilities progress roughly doubled between October 2025 and October 2026, and that AI automating AI R&D is probably the large majority of the reason. The author argues this means the "dramatic acceleration" threshold in Anthropic's Responsible Scaling Policy has already been crossed. Because researchers are only one input to progress, sustaining that pace implies a labor uplift of perhaps 10x — matching informal staff reports of ~12x.
It's as if every researcher at Anthropic is in a time warp bubble where they experience ten days for every day that passes in the outside world.