[METR: Anthropic Model Progresses Approximately Twice as Fast as a Year Ago] According to a report by The Verge on November 7th, AI security research firm METR, after an 11-week investigation of Anthropic, stated that the company's models are progressing at approximately twice the rate of a year ago, with the progress achieved in the past year equivalent to about two years previously. METR believes that as of October, the majority of the accelerated R&D was due to AI, and its contribution is growing. Currently, the acceleration primarily stems from AI assisting human researchers, rather than autonomous work. Employees using Anthropic's most powerful internal models reported in informal surveys that their productivity was, on average, about 12 times higher than without AI assistance. The report notes that Anthropic heavily redacted its public reports. Co-founder Sam McCandlish stated that the information contained commercially sensitive content and could potentially help other companies accelerate their R&D.