The BaseEssayAI Alignment Forum

Q3.5 Timelines Update: Looks Like the Intelligence Explosion Will Happen in 2027 After All

In a linkpost from the AI Futures Project, Thomas Larsen and co-authors take METR's estimate of AI-driven acceleration at Anthropic as input to their models. They now forecast a Superhuman AI Researcher likely in 2027, call for a pause or pacing of frontier AI, and criticise Anthropic's redactions.

Open the source ↗3 quotations checked, word for word ↓Top story of the week
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

The authors argue that AI systems doing AI research themselves could arrive within about a year, which they say is faster than they expected only three months ago.

For lawmakers

The AI Futures Project authors urge an industry-wide pause followed by an international agreement, and ask Anthropic to release enough information for outsiders to assess METR's acceleration estimate.

For researchers

The authors filter their AI Futures Model on METR's acceleration estimates and list candidate reasons the model was wrong, but say their analysis is highly preliminary.

The question it raises

If an AI could replace the best human researchers, the authors argue, the question of who steers its development becomes urgent before that point is reached.

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.

“We estimate that METR’s report implies a 1.7x AI-attributable acceleration in the pace of progress today, relative to historical levels.”
Thomas LarsenThe authors state their reading of METR's assessment of AI R&D acceleration at Anthropic, the input to their updated forecasts.
Found word for word in the source
“If SAR is coming in 2027, it is incredibly urgent to pause or substantially ‘pace’ frontier AI development, with the ideal being an industry-wide pause followed by an international agreement.”
Thomas LarsenThe authors draw a policy conclusion from their forecast.
Found word for word in the source
“We aren’t sure what reasons Anthropic gave internally for redacting all of this important information, but we are very skeptical that Anthropic is fairly assessing the costs and benefits from an impartial perspective.”
Thomas LarsenThe authors comment on Anthropic's redaction of much of the evidence in METR's report.
Found word for word in the source

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
Thomas Larsen (Wrote it) · Daniel Kokotajlo (Wrote it)
What we read
We could read only part of the source.
Found by
The Radar, run of November 11, 2026, via ai-alignment-forum (tier 1: the source’s own channel or an official record)
Cite this
AI Alignment Forum, “Q3.5 Timelines Update: Looks Like the Intelligence Explosion Will Happen in 2027 After All”, November 10, 2026. Entry in the Ctrl AI Base, https://ctrlai.com/base/alignmentforum-2026-11-10-q3-5-timelines-update-linkpost
Our notesRead the first part (summary, METR section, start of modelling) of a 64-minute linkpost; the later sections were not read. The Base already holds an entry on the same post from the AI Futures Project's own blog (alignmentforum-2026-11-10-aifp-intelligence-explosion-2027); this one is the Alignment Forum copy and may be a duplicate to merge. Byline lists six authors; only Larsen and Kokotajlo are named as people here, the other handles are not full names. Concerns Anthropic, so conflict is true.