Why METR’s Report on Anthropic’s AI Research Acceleration Alarms Researchers

What happened

The Verge examines a METR report based on an 11-week investigation that found Anthropic’s models were improving roughly twice as quickly as a year earlier, with researchers attributing most of the acceleration as of October to AI. METR said employees using the company’s most capable internal model reported being around 12 times as productive, but noted that Anthropic redacted much of the evidence in the public report. Researchers cited in the article raised concerns about the pace of progress and the limits on independent scrutiny.

Read next

• More changes in today’s briefingMETR Says Anthropic May Have Crossed Its AI Acceleration Warning Threshold in July

The story so far

METR reported that Anthropic’s AI research and development accelerated, with researchers attributing most of the gains measured as of October to AI. The assessment followed an 11-week investigation inside the company and found that Anthropic’s models were improving roughly twice as quickly as a year earlier. METR said employees using Anthropic’s most capable internal model reported being around 12 times as productive. Researchers cited by The Verge raised concerns about the pace of progress and independent scrutiny; the article said Anthropic redacted much of the evidence in METR’s public report.