Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Clear out junk files and repair common Windows errors3Fix the driver behind crashes, sound loss and screen glitchesiTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
According to a new arXiv critique, parts of OpenAI’s Lean formalization do not faithfully match the natural-language proof released alongside its claimed Navier–Stokes result. The paper points to specific differences, including an estimate that appears to require an additional input derivative in Lean and a pressure-flux estimate whose formal statement and argument differ from the written version. Those findings raise a translation and verification issue; they do not, by themselves, settle whether the overall mathematical result is correct.
What the critique says was mistranslated
OpenAI says an internal system produced a proof that solutions to the Navier–Stokes equations can develop a singularity in finite time, and that it shared both a written proof and a Lean formalization. The paper “Navier-Stokes lost in translation” examines whether the formal development corresponds to claims in the prose proof. Its authors report mismatches in examples they analyze; they do not present those examples as a complete independent audit of every line of code.
The derivative estimate
In comparing the natural-language estimate in Lemma 8.6 with cited Lean declarations, the paper’s authors say the written result claims control with one fewer input derivative than the formalized estimate appears to require. They describe the difference as an m+4 versus m+5 derivative requirement. This is the authors’ technical reading of the proof and code, not a standalone determination of the whole result.
The pressure-flux bound
The authors also compare a pressure-flux bound and its proof in the prose with the corresponding Lean estimate and formal argument. They say these differ, including because the Lean estimate depends on an additional quantity absent from the written bound. The issue is not simply whether Lean accepted a proof: it is whether the theorem and argument encoded in Lean are the same ones the written proof claims.
#1 Best Overall
What Lean verification does—and does not—establish
Lean is a proof assistant: it checks that a formal statement follows from definitions and proof steps in its formal environment. If Lean accepts a proof, that is evidence about the encoded theorem. It does not automatically confirm that a mathematician translated a separate prose theorem into the formal statement faithfully. That correspondence must be checked independently, which is the central distinction in the paper’s critique.
There are four separate questions in this episode, and evidence about one does not answer all the others:
| Question | What it asks |
|---|---|
| Is the formal Lean theorem proved? | Whether the proof assistant verifies the theorem stated in its formal environment. |
| Does the formal theorem match the prose theorem? | Whether the code encodes the claims made in the written proof. The arXiv authors identify mismatches in the examples they discuss. |
| Is the natural-language proof valid? | Whether its mathematical reasoning establishes its stated result. A discrepancy between prose and code does not by itself resolve this. |
| Does the result answer the intended problem? | Whether the mathematical setting is the version of the Navier–Stokes problem readers and experts want answered. This is a distinct scope debate. |
Has the entire proof been checked by humans?
Science News reported that mathematicians were still digesting the long proof at the time of its coverage. It quoted Johns Hopkins mathematical physicist Gregory Eyink: “I don’t think anyone has completely verified the proof yet, certainly not on the human side.” That quotation describes the state of review when Science News reported; it should not be treated as a current, timeless status update or as a judgment on the specific discrepancies in the arXiv paper.
Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallOutdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchA technical critique documenting mismatches is important evidence, but it is not equivalent to a completed independent verdict on every part of the formalization and written proof. The available reporting does not establish that the entire proof has since been fully verified by human experts.
Is this also a dispute about the right Navier–Stokes problem?
Yes, but it is a separate dispute. Scientific American reported criticism that the result may concern a variant of the problem that some experts regard as disconnected from physical reality or less interesting. That concerns the scope and significance of the mathematical setting, not whether the Lean code faithfully represents the prose. The two questions should not be conflated: a faithful formalization would not, on its own, settle whether the chosen setting answers the intended problem, and a scope objection would not establish a translation error. See Scientific American’s reporting on the problem’s scope.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What OpenAI says about the work’s origins
In its announcement, OpenAI says the work began after it heard a rumor it later connected to Tristan Buckmaster and Levent Alpöge. OpenAI characterizes their result as concerning forced Euler, says it offered them access to its prompts and later proof, and recognizes their priority on forced Euler. These are OpenAI’s account and characterization; the reviewed sources do not independently resolve every question about priority or access to data.
Rank #4
OpenAI also described the group working on its Navier–Stokes result as involving on the order of 10,000 concurrent agents and said it does not intend to claim the Millennium Prize for the result. Those are statements about the company’s process and position, not independent verification of the proof’s correctness.
Quick wins for a faster PC:
Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Clear out junk files and repair common Windows errorsFree Scan →Scan for outdated or missing drivers - takes under a minuteDriver Scan →Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
