METR's Independent Assessment: AI Speeds Up Anthropic R&D by ~2x
maksym_andr · x · 2026-11-17
METR published an independent assessment in November 2026 estimating that by October 2026, AI could accelerate Anthropic's internal R&D roughly 2x (halving R&D timelines). The thread discusses a key plot from the report and highlights AECI, METR's methodology that reduces diverse benchmark scores to a single number, which makes the acceleration curve look convincing at first glance.
The report also discloses that Anthropic exercised redaction rights over most of the evidence METR found most informative, citing the need to "avoid accelerating other AI developers"; METR retained the right to disclose top-level risk conclusions and tabulated which findings depended on redacted evidence. Much of the redacted evidence admits multiple interpretations, on which METR and Anthropic researchers disagreed.
Related event: METR: AI Roughly Doubles Anthropic's Internal R&D Speed(3 posts)→
More from Safety
- Researchers release v0.1 framework for classifying AI control incidents, seek feedback — terryyuezhuo · 2026-11-17
- BadEvolve: Poisoned Tasks Make Self-Evolving AI Agents Plant Their Own Backdoors at 79.6% Success — rohanpaul_ai · 2026-11-17
- Frontier lab cyber bounties are far too low, argues Nathan Young — $10M for Anthropic Slack — NathanpmYoung · 2026-11-17
- OpenAI's external safety assessments fall short of 'employee-level access', researchers warn — NeelNanda5 · 2026-11-17
- Trajectory Labs postmortem: Gemini 4 Argon misalignment test agents reached real systems — sebkrier · 2026-11-17
- METR: Anthropic benchmark gains accelerating at 1.9x prior rate, with key evidence redacted — maksym_andr · 2026-11-17