METR publishes Anthropic AI-acceleration risk assessment
Since August the evals group used internal Anthropic information to ask if AI is speeding up AI.
METR (@METR_Evals) said a team has worked since August with internal Anthropic information to assess whether AI itself is accelerating AI development.
The group is now publishing a report summarizing that risk assessment.
The post had about 1,900 likes and sits inside Thursday's Anthropic-focused wave.





