The BaseEssayDon't Worry About the Vase

AI R&D Is Accelerating Towards Potential RSI in 2027

Zvi Mowshowitz reads METR's report on AI-driven acceleration of AI research at Anthropic, which estimates about twice the earlier rate of capability progress. Mowshowitz argues this points to recursive self-improvement within about a year, criticises Anthropic's redactions and urges Congress to subpoena Anthropic.

Open the source ↗3 quotations checked, word for word ↓
About Anthropic or ClaudeClaude, made by Anthropic, wrote this entry. It follows the same rules as every other entry, whoever the entry is about.

Why it matters

The same thing, for different readers.

What it means for you

An independent evaluator says Anthropic's AI progress runs at about twice its earlier pace, and one commentator says that could lead to systems that improve themselves within a year.

For lawmakers

Mowshowitz asks Congress to demand the full METR report and what Anthropic's leaders know, since the public version has most of its strongest evidence redacted.

For researchers

METR says Anthropic redacted a majority of the evidence it found most informative, so the public must rely on METR's judgment more than METR considers ideal.

The question it raises

If AI helps build better AI faster than people can assess it, the question is who gets to decide when to stop.

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.

“A naive extrapolation says we may be less than a year away from full RSI. There are no clear bottlenecks in the way.”
Zvi Mowshowitz, Writes Don't Worry About the VaseSummarising what METR's findings imply, in the opening of the post.
Found word for word in the source
“I would also suggest to Congress that they should want to see the full report along with everything else Anthropic and Dario Amodei know and really think about this, and to do that they should follow Kelsey Piper’s suggestion and subpoena Anthropic.”
Zvi Mowshowitz, Writes Don't Worry About the VaseAfter describing Anthropic's redactions of the METR report.
Found word for word in the source
“As of October 2026, Anthropic is likely improving the capabilities of their models on well-specified tasks at around twice the rate they did between March 2024 and October 2025.”
METRMETR's stated conclusion, quoted in Zvi Mowshowitz's post; METR adds that the estimate carries high uncertainty.
Found word for word in the sourceAs quoted in the text we read, not the speaker’s own text or a transcript of their wordsstarts inside a longer claim, right after “that”

The record

Questions
What happens when AI helps build more capable AI? · How fast is AI advancing, and where is it heading? · Who can demand evidence, delay a release or stop an unsafe deployment?
People
Zvi Mowshowitz (Wrote it) · Ajeya Cotra (Mentioned, METR) · Kelsey Piper (Mentioned) · Dario Amodei (Mentioned)
What we read
We could read only part of the source.
Found by
The Radar, run of November 10, 2026, via zvi-mowshowitz (tier 1: the source’s own channel or an official record)
Cite this
Don't Worry About the Vase, “AI R&D Is Accelerating Towards Potential RSI in 2027”, November 9, 2026. Entry in the Ctrl AI Base, https://ctrlai.com/base/zvi-2026-11-09-ai-rd-accelerating-rsi-2027
Our notesOnly the first part of the long post (about 180 of its lines, up to the section on the pace of progress) was read for this entry. The METR quote is METR's text as reproduced by Zvi; the original METR report is a separate entry.