argumentA line of reasoning from premises to a conclusion. Like a claim, it belongs to no one who makes it.constitution →

The argument from METR's attribution estimate

forThis argument, if it holds, bears in favour of the claim.constitution →AI-assisted research and development caused most of Anthropic's capability acceleration since October 2025.

The argument

METR, after an eleven-week high-access assessment, estimated that roughly 80% of Anthropic's acceleration as of October 2026 is attributable to AI assistance, ruled out faster training-compute scaling as the cause, and judged that the other factors it examined would not explain a sustained acceleration. Staff interviews centred on about a twofold AI speedup. METR's judgment is reliable on this question, and AI's share grew over time. So AI assistance plausibly accounts for most of the acceleration over the whole period.

Does it work?

Does it follow?: slight even if the premises holdEven if the premises are true, they make the conclusion only a little more likely.constitution →
Do the premises hold?: 1 not separately assessed

Granting its premises, the argument shows that AI assistance is the main cause of the acceleration in late 2026. It does not show the same for the whole period since October 2025, because the instantaneous share (linked claim) is expected to exceed the cumulative share, by an amount METR's redacted estimates would state. It also rests on METR's reliability on evidence the public cannot see. On the claim as worded it moves the case only somewhat.

Premises

UnassessedNo current assessment. Attention goes where its expected value is highest and someone funds it; nothing has funded an assessment of this claim yet, and anyone can.constitution →Share of Anthropic's observed AI-capability acceleration in October 2026 attributable to AI assistance (percent)

Where it is made

Passages where sources make, criticize or mention this argument. They record where it appears, not how strong it is, and none has precedence over another.

Independent Assessment of AI R&D Acceleration at Anthropic in October 2026mentions it
We understand that AI is not yet able to make highly significant and fully autonomous contributions to the most important parts of the model training process. However, researchers described significant “signs of life” implying such contributions could occur soon, although with substantial variation depending on the area. Further details were redacted. This evidence came from: