The BaseEssayAI Alignment Forum
Ajeya: AI is dramatically accelerating AI research right now
Zach Stein-Perlman's linkpost on the AI Alignment Forum reproduces the opening of an essay by Ajeya Cotra. It argues that METR's report on Anthropic, read against Anthropic's own Responsible Scaling Policy, shows the "dramatic acceleration" threshold has already been crossed.
Why it matters
The same thing, for different readers.
What it means for you
Ajeya Cotra argues that AI is already roughly doubling how fast Anthropic's AI capabilities improve, which would bring powerful AI sooner than most outside forecasts assumed.
For lawmakers
Ajeya Cotra reads Anthropic's Responsible Scaling Policy as already triggered by METR's findings, which raises the question of what Anthropic and regulators now owe the public.
For researchers
METR's measure of about twice the earlier pace rests on an internal index, and Ajeya Cotra notes that most of the supporting evidence was redacted, so outsiders cannot check it.
The question it raises
If AI is doubling overall progress, Ajeya Cotra argues, it must be speeding up human researchers by much more, which raises who stays in control of the loop.
In their own words
Every quotation was found, word for word, in the text we saved from the source. That shows the words are there; it does not make them true, or prove the speaker meant them as quoted.
From the source’s own text.
“From my reading of their report and Anthropic’s RSP, it looks like the “dramatic acceleration” threshold has already been crossed.”
“Between October 2025 and October 2026, capabilities progress at Anthropic has been going about twice as fast as it had been over the previous twelve months.”
“I think this report is the most important update we have gotten about AI risk in the last couple of years — much bigger than the Hugging Face attack.”
How the finding was reached
In our words, from the source. Weigh the finding by how it was made.
- What was done
- A deep-access investigation by METR staff of how much Anthropic's AI systems accelerate its own capabilities progress, measured with an internal version of the Epoch Capabilities Index (AECI).
- What it found
- Capabilities progress at Anthropic ran about twice as fast between October 2025 and October 2026 as over the previous twelve months, as relayed by Ajeya Cotra.
- What it does not show
- Ajeya Cotra says a wide range of supporting evidence was redacted, including the reasons for believing AI automation is most of the cause. This entry has only her summary, not the report.
- Status
- A report
The record
- Questions
- What happens when AI helps build more capable AI? · How fast is AI advancing, and where is it heading?
- People
- Ajeya Cotra (Speaks) · Zach Stein-Perlman (Wrote it)
- What we read
- We could read only part of the source.
- Found by
- The Radar, run of November 8, 2026, via
ai-alignment-forum(tier 1: the source’s own channel or an official record) - Cite this
- AI Alignment Forum, “Ajeya: AI is dramatically accelerating AI research right now”, November 8, 2026. Entry in the Ctrl AI Base, https://ctrlai.com/base/alignmentforum-2026-11-08-ajeya-ai-accelerating-ai-research