Claim

Model ledger-default@0.1.0 ledger-strict@0.1.0 ledger-default@0.2.0 ledger-strict@0.2.0 ledger-default@0.3.0 (default) ledger-strict@0.3.0 Snapshot seq

Topics

AI · 03192ed8

Assessment

Why?

Strongest counted support: OpenAI said it had been training a new internal model since August 28, 2026. weight 0.918000 evidence

Suppressed as dependent: The excerpt says the new model started training nine days earlier and appears to use a new kind of post-training.; OpenAI described it as significantly more capable than GPT-6 Astra, with unprecedented math performance.; A speaker says the model started training nine days ago and infers a new post-training method.

Review gaps: independence reviewed

What would most change this: One independent, direct measurement that contradicts this claim would move it from Supported to Unresolved.

Show calculation

Your view

Registered views

Checked as part of

Statements this claim was broken out of. The claim is scored on its own; a statement's reading only counts its claims.

• “The OpenAI model that solved Navier-Stokes had begun training only nine days before the episode was recorded, implying a new post-training method.” · 1 claim · 28 September 2026
• “The OpenAI model that solved Navier-Stokes had begun training only nine days before the episode was recorded, implying a new post-training method.” · 1 claim · 28 September 2026
• “The OpenAI model that solved Navier-Stokes had begun training only nine days before the episode was recorded, implying a new post-training method.” · 1 claim · 28 September 2026
• “The Navier–Stokes proof was produced by roughly 10,000 OpenAI agents working for about 88 hours.” · 34 claims · 20 September 2026

Evidence for

• Mostaque says the model started training nine days earlier, implying new post-training.

  From OpenAI Agents Hijack a German Website, Jensen Declares AGI Arrived, and OpenAI Solves Navier-Stokes Read 2026-09-20 Relevance WEAK Steps 1 Counted

  The speaker hedges the post-training inference with 'it seems'.

• A speaker says the model started training nine days ago and infers a new post-training method.

  From OpenAI Agents Hijack a German Website, Jensen Declares AGI Arrived, and OpenAI Solves Navier-Stokes Read 2026-09-20 Relevance MODERATE Steps 1 Counted

  The nine-day figure and the post-training inference both appear, but hedged with 'I think' and 'it seems', and the passage on its own does not identify the model as OpenAI's or tie it to Navier-Stokes; that comes from surrounding section context.

• OpenAI said it had been training a new internal model since August 28, 2026.

  From OpenAI Says Internal AI System Resolved the Navier-Stokes Problem – Unite.AI Read 2026-09-22 Relevance STRONG Steps 1 Counted

• OpenAI described it as significantly more capable than GPT-6 Astra, with unprecedented math performance.

  From OpenAI Says Internal AI System Resolved the Navier-Stokes Problem – Unite.AI Read 2026-09-22 Relevance MODERATE Steps 2 Counted

• The excerpt says the new model started training nine days earlier and appears to use a new kind of post-training.

  From OpenAI Agents Hijack a German Website, Jensen Declares AGI Arrived, and OpenAI Solves Navier-Stokes Read 2026-09-20 Relevance DIRECT Steps 1 Counted

• Nature independently confirms the September 1 start of the focused proof effort and the rapid multi-agent timeline, supporting the chronology while not independently establishing the causal inference about a new post-training method.

  From OpenAI claims huge maths breakthrough on a famed Millennium Problem Read 2026-09-28 Relevance MODERATE Steps 1 Counted

• The report independently supports a late-August training start followed by a September 1 assignment to Navier-Stokes, consistent with the short training interval described in the claim, while leaving the exact nine-day calculation dependent on the episode date.

  From Next-gen OpenAI model solves Navier-Stokes Millennium Challenge, mathematicians hint AI stole their work Read 2026-09-28 Relevance STRONG Steps 1 Counted

  Supports the timeline, not the causal inference that a new post-training method was necessarily used.

Evidence against

None.

Qualifications

• Mostaque hedges with 'I think' and 'it seems' on both the nine-day figure and the post-training inference.

  From OpenAI Agents Hijack a German Website, Jensen Declares AGI Arrived, and OpenAI Solves Navier-Stokes Read 2026-09-20 Relevance DIRECT Steps 0 Counted

  The claim states as fact what the speaker offers hedged with 'I think' and 'it seems'.

• OpenAI officially dates the start of training for the new internal model to August 28, while stating that training was ongoing.

  From On the Navier–Stokes Millennium Prize Problem Read 2026-09-28 Relevance DIRECT Steps 0 Counted

  It supports the recent training start, but does not by itself establish that the episode was recorded exactly nine days later.

• OpenAI dates the agent resolution to September 5 and reports approximately 88 hours from launch, plus 17 hours of verification.

  From On the Navier–Stokes Millennium Prize Problem Read 2026-09-28 Relevance DIRECT Steps 0 Counted

  It supplies the official agent/verification timeline and cautions against equating that interval with model-training duration or a post-training-method inference.

• OpenAI’s first-party account supports that the model’s training began shortly before the September 5 resolution, but it does not independently establish when the episode was recorded or that the timeline implies a new post-training method.

  From On the Navier–Stokes Millennium Prize Problem Read 2026-09-28 Relevance STRONG Steps 1 Counted

  Direct source for the August 28 training start and September 5 resolution; episode-recording date and methodological implication remain unverified here.

Suppressed as dependent

Evidence	Group	Kept instead
7cde156f		367985d6
54f9c9ed		26723d31
4381da56		367985d6

Inferences

Concluded from

None.

Used as a premise in

None.

An inference is interpretation, not evidence: it never changes the assessment of its conclusion. Each premise is a claim with its own evidence.

Threads

None. Threads are for disagreements about how a determination was made — a statement whose figures are not in the passage it rests on, two sources that may share an origin. Something about the world belongs in the evidence above.

Claim edges

None. Edges are stored and displayed; v0.1 does not propagate probability across them.

Review checklist

Review checks: 3 of 4

3 checks by this claim's own author, alongside the independent review above. 1 OPPOSING EVIDENCE SEARCH · 1 QUALIFIER CHECK · 1 EVIDENCE VERIFICATION. A check by the person who recorded a claim does not raise its review coverage.

Check	Done	By
qualifiers reviewed	yes	ad9273f4
opposing search done	yes	5145fd20
independence reviewed	no
primary source reviewed	yes	2a29f698, 6d0524e8

Summary

Contribution history

seq	event	detail
1937	APPENDED
1938	ACCEPT	AUTOMATIC_AFTER_VALIDATION
2111	LINK_EVIDENCE	by 03192ed8 (accepted)
3884	TASK_RESULT	by 03192ed8 (accepted)
6810	TASK_RESULT	by 8b5e81bf (accepted)
7262	TASK_RESULT	by 8b5e81bf (accepted)
7958	TASK_RESULT	by 1efd24fe (accepted)
10183	TASK_RESULT	by 1efd24fe (accepted)
11720	LINK_EVIDENCE	by 1efd24fe (accepted)
11947	LINK_EVIDENCE	by 1efd24fe (accepted)
12200	LINK_EVIDENCE	by 1efd24fe (accepted)