Unforced Navier-Stokes Singularity And The Generalized Riemann Hypothesis
AIThis post was created with the assistance of artificial intelligence (AI).

TL;DR

Terence Tao reported three AI-assisted proof efforts: finite-time blowup for unforced 3D Euler, finite-time blowup for unforced 3D Navier-Stokes, and a Lean-only proof of the Generalized Riemann Hypothesis. He described the Navier-Stokes strategy as broadly clear but said the proof needs better exposition; verification and understanding of the GRH proof are at an early stage.

Mathematician Terence Tao reported on November 20 that researchers have produced AI-assisted, Lean-formalized proofs claiming finite-time blowup for the unforced three-dimensional Navier-Stokes equations, alongside a separate Lean-only proof claiming the Generalized Riemann Hypothesis. Tao presented the results as at different stages of scrutiny, not as settled mathematical conclusions: the Navier-Stokes argument still needs professional exposition, while the GRH proof is enormous and its strategy is only beginning to be examined.

Age 18–24?Offer from Amazon

Prime made for students and young adults

  • Fast, free delivery for dorm and study essentials
  • Prime Video and Amazon Music included
  • Member-only deals
Try Prime for Young Adults Free trial for eligible 18–24 year olds
As an affiliate, we earn on qualifying purchases.

Tao said Levent Alpöge, Tristan Buckmaster and Matei Coiculescu wrote a human-readable proof of finite-time blowup for the unforced incompressible Euler equations in three dimensions, with heavy AI assistance, and formalized it in Lean. He also reported that Alpöge and Buckmaster generated a proof making the corresponding blowup claim for the unforced three-dimensional Navier-Stokes equations, again with substantial AI use and a Lean formalization.

For Navier-Stokes, Tao said the paper is technically human-readable and that its broad strategy is clear. It constructs a numerical approximate discretely self-similar solution, then applies rigorous perturbation theory to seek an exact solution. Tao said more work is needed to turn the material into a properly written proof. The report does not supply a journal publication or independent validation of the result.

Tao separately said Alpöge used an internal AI model to generate an enormous Lean-only proof of the Generalized Riemann Hypothesis (GRH). Initial analysis, he wrote, suggested the proof draws on modern developments in the Langlands program and arithmetic geometry. Tao directed readers to a Lean Zulip discussion for verification efforts and attempts to extract the proof’s strategy; moderators later narrowed that discussion to Lean-related matters.

At a glance
reportWhen: Reported November 20, 2026; verificatio…
The developmentTerence Tao published an update describing three AI-assisted mathematical proof claims, including a Lean-formalized Navier-Stokes blowup result and a Lean-only proof of the Generalized Riemann Hypothesis.

Three Claims, Different Verification Stages

The claims concern major open problems in partial differential equations and number theory. A valid finite-time blowup proof for three-dimensional Navier-Stokes would establish that smooth solutions can develop a singularity in finite time under the conditions stated. A proof of GRH would resolve a central conjecture about the distribution of prime numbers and related questions in number theory.

For now, the immediate news is not that these problems have been definitively solved, but that researchers report producing formalized proof artifacts with heavy AI assistance. Lean formalization can help check that a proof follows from its definitions and accepted foundations once the relevant code is correctly stated and the system and dependencies are trusted. It does not by itself make a result easy for mathematicians to understand, nor does Tao’s account establish that the proofs have completed independent review. The gap between a machine-checkable artifact and a clear, scrutinized mathematical argument is especially visible in the GRH claim.

As an affiliate, we earn on qualifying purchases.

How Tao Ranked the Proof Efforts

Tao introduced the results in a post titled “Unforced Euler Singularity, Unforced Navier-Stokes Singularity, and the Generalized Riemann Hypothesis.” He said he was listing them in decreasing order of how “digested” they were at the time. In his account, the Euler result had a human-readable proof and Lean formalization; the Navier-Stokes work had a readable paper and an identifiable broad approach, but needed stronger exposition; the GRH work was a large Lean-only proof whose strategy remained to be extracted.

The distinction matters because the two fluid equations are related but not interchangeable. The reported Navier-Stokes claim concerns the unforced incompressible equations in three dimensions; it should not be broadened to other settings or read as a resolution of every regularity question. The GRH statement is a separate number-theory claim. Tao also pointed to a Mastodon post by Buckmaster for information on the first two results, and said a human author was available to help explain the GRH proof.

Verification and Proof Details Remain Open

Tao’s report does not say that independent experts have completed a full verification of any of the three results, or that the claims have been accepted through peer review. The Navier-Stokes proof’s full written presentation is still a stated need, and the source gives no detailed account of the formalization, reviewers, or timeline for that work.

For GRH, the proof’s size and Lean-only form leave its underlying mathematical argument difficult to assess from the report alone. Tao described the initial reading of its methods as tentative. It is also unclear how much of the reported AI work involved generating the mathematical ideas, translating existing reasoning into formal code, or other forms of assistance. The source does not provide enough detail to establish those roles or to independently confirm the claims.

Reviewers Must Explain and Check

The next steps described by Tao are continued examination of the Lean proof artifacts and efforts to recover and communicate their mathematical arguments. The Lean Zulip thread is a place to follow the GRH verification work, although moderators have limited its discussion to Lean-related topics. Tao also indicated that Alpöge could help explain the proof. For Navier-Stokes, the reported need is a more polished, professionally written proof that specialists can scrutinize.

No deadline for completion or formal publication was given. Until more detailed proofs and independent assessments are available, the claims should be treated as reported proof efforts under review, rather than established resolutions of the Euler, Navier-Stokes or GRH problems.

Key Questions

Has the Navier-Stokes problem been solved?

Tao reported a proof effort claiming finite-time blowup for the unforced incompressible equations in three dimensions. He said the broad strategy was clear but that the paper still needed professional exposition; his post does not establish that the result has completed independent verification.

What is the reported Navier-Stokes proof strategy?

According to Tao, it builds a numerical approximate discretely self-similar solution and uses rigorous perturbation theory to convert that approximation into an exact solution.

What is the GRH proof claim?

Tao said Levent Alpöge generated an enormous Lean-only proof claiming the Generalized Riemann Hypothesis, using an internal AI model. Tao reported that efforts to verify it and understand its strategy were at an early stage.

Does Lean formalization mean the proofs are independently accepted?

No. Lean can check formalized reasoning within its system and dependencies, but a formalization is not the same as completed independent mathematical review. Tao’s post describes verification and exposition work still underway.

Source: terrytao.wordpress.com

You May Also Like

How to Tell When a Piece Needs Better Environmental Control

Only by recognizing signs like peeling, mold, or pests can you know when your collection needs better environmental control to prevent damage.

Biggest Dark Matter Detector Spots A Single Weird Particle

The world’s largest dark matter detector has detected a single anomalous particle, sparking scientific intrigue and further investigation.

WeatherNext 3

DeepMind’s WeatherNext 3 aims to revolutionize weather forecasting with enhanced accuracy, but details remain preliminary and unconfirmed.

Open Thread 449

Search interest in Open Thread 449 is surging, driven by a spike in online activity. Exact reasons remain unconfirmed, but trends suggest heightened engagement.