Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
[flagged] OpenAI's claimed disproof of Connes' Rigidity Conjecture is invalid [pdf] (philarchive.org)
31 points by muglug 9 hours ago | hide | past | favorite | 42 comments
 help



The author of this paper has also apparently published a proof of the Riemann Hypothesis. yeah idk if they should be trusted as an authority on this.


Crank author at a crank “institution” - I can’t assess whether the OpenAI paper is accurate but I highly doubt this is going to be the paper to disprove it.

University of Kansas is not a crank institution. Author has MS and MA from KU and BSc summa cum laude / Valedictorian of her departments from UMKC. Trying to call her a crank for hosting a private think tank is jealous nonsense.

Author is a crackpot. She does not meaningfully engage with anyone who points out the key flaw in her counterargument. See the thread here https://x.com/AcerFur/status/2083649346294382803

Author is a PhD candidate who graduated summa cum laude in BSc Physics Valedictorian of her college and math with double master's degrees. You're the crackpot clearly, you provide no correct disproof of Nielsen.

This attack on Nielsen's paper is crackpottery / academic revenge porn by John Synons. To base an attack on Nielsen on frickin Acerfur is using an undergrad who hasn't passed his first semester yet to take on a PhD student who graduated in physics and math summa cum laude. Nonsense.

Thanks — sorry for spreading nonsense.

acerfur is an undergrad with zero experience. OpenAI hired him for slave labor. Can't possibly perform at this level, he has no background. Zip.

Get help.

The nonsense is the crackpot who wrote this evaluation. Nielsen is an established physicist close to completing two PhDs.

This is complete nonsense accusation. She disproved Acerfur completely.

Genuine question - this is so hard for me to follow. Not that I could follow the original disproof anyways. But how is "truth" determined when the effort required to validate is so high?

It's a common problem in math. The famous proof for Fermat's last theorem took 2 years to validate.

That's different though. Understanding the spec of Fermat's last theorem is simple. Here, you already need to know some mathematics just to understand the spec.

Nielsen's proof is correct and verified by three AIs against OpenAI now.

Not sure what your comment has anything to do with my comment, or the comment I replied to.

But to answer to your comment: Opus 5 says that Nielsen's critique is nonsense.



The crackpot is the guy posting here. He's an ex of Nielsen who constantly bashes her in economics forums to try to throw off the google algorithm. it's basically the former Philosopher Chair John SYmon's equivalent of revenge porn, but they never even dated. He hates her for dumping him for a more physics oriented approach and constantly harassses her. See: https://www.youtube.com/watch?v=hWQnZWANIwA

These AI generated proofs have something akin to the quantum algorithms that can generate a response that would take 1 billion years of processing to finish on a classical computer. How do you test them to see if the result was correct?

No they don't. They have a few thousand lines and the hinge points are easy to find. This is more nonsense.

Maybe there's an error, but much of this write-up reads like nonsense to me. The assertion that a semidirect product with an abelian factor must admit that factor in its center is absolutely false. This is actually acknowledged later in this article, but is handwaved away in incomprehensible fashion.

You're not a mathematician and have no business being involved in this at all. The author the OP called a "Crackpot" graduated Valedictorian of University of Missouri at Kansas City. She provides detailed proofs in lean with correct labels. Grok, Claude, and Gemini have verified her proofs. Open AI Sol and above will not attempt to disprove a standing result from Open AI it's programmed not to and admits that, Nielsen showed it admitting that. OpenAI also admitted NIelsen's result is serious and merits consideration, despite the fact it's programmed not to acknowledge disproofs of OpenAI materials.

Nielsen is an expert, not a crank. She worked at I2S at the University of Kansas and passed her master's and her PhD thesis in PhilSci / PhiLMath is on the web and completely airtight.

This "paper" itself is 100% AI-generated...

What makes you say that? Not dunking, I skimmed it (I’m in no way at this level) and didn’t see anything outright Claud-y.

This immediately struck me as Claud-y:

> The gap has two independent consequences, each sufficient to invalidate the claimed disproof. The first is structural.

Then the classic coding agents negation of earlier evidence, instaed of just updating to use new references, they mention that they changed old to new:

> The publicly released monolithic file ConnesRigidity.lean (37,000+ lines) does not use the names CocycleExtension, ZeroCocycle, or TwistedCocycle that appeared in the earlier modular source files (CocycleExtension.lean, ICC.lean, CrossedClosure.lean). However, the identical mathematical construction is present under different names. The following table gives the correspondence, with line numbers in the published file.

> This is the same zero-cocycle / twisted-cocycle structure identified in the earlier modular source files, confirming that the structural analysis of this note applies to the published code

And some other Claud-y stuff:

> Why both paths are closed. A successful defence would have to close both paths simultaneously

> This case illustrates a failure mode that is becoming increasingly well documented in the literature on AI-assisted formal mathematics: the gap between what a formal proof verifies and what it means. The Lean kernel certifies that a proof term inhabits a given type; it does not certify that the type faithfully encodes the intended mathematical claim. As Tao has emphasised

And afterwards I cross-verified with Pangram 4 which I trust, it marked the preamble/starting stuff as 100% AI-generated.


You're using the wrong names from an older version of their paper and Nielsen's , she's updated to include the new relabels. The proof holds.

That's false. It's not AI generated. It's AI checked. Nielsen has a style that often flags as AI because she uses many EM dashes and has a very rigorous style. Her papers from 2015-2020 also flag as AI generated, before AI was available.

Can you provide a single example of such paper that would show as AI generated on Pangram?

This attack on Nielsen's paper is crackpottery / academic revenge porn by John Synons. To base an attack on Nielsen on frickin Acerfur is using an undergrad who hasn't passed his first semester yet to take on a PhD student who graduated in physics and math summa cum laude. Nonsense.

A lot of people who know mathematics say Nielsen’s paper is riddled with errors. I regret sharing it here.

> Undergrad who hasn’t passed his first semester yet

Not sure what you’re talking about. According to his wiki bio he’s done two years at Cambridge. He posted Maths questions from the 1B tripos exam, and was smart enough to get an OpenAI internship.

Whereas it seems like the author of this paper has never been hired anywhere by any company or institution — the "Center for Topological Physics" listed seems to exist only as a website.


One of mathematicians working at OpenAI refuted those claims directly on X - https://x.com/AcerFur/status/2083656978719719601

Yes that guy is an undergrad with ZERO math background. HAsn't even finished his first semester. He's very proud of being a wage slave with zero experience. He can't take on someone with a BSc Valedictorian of a large urban university with two master's degreees damn near to two PHDs. That's very unfair to him.

Good grief that thread is sad, with it ending by Gary Marcus asking for the crank to be taken seriously.

So this is the level of science discourse now? AI really broke people's minds. If it's true or not, it will be eventually proved or not. Name calling really is the best you all can do?

does this show Lean is not bulletproof?

Lean never was bullet proof. Nielsen also provides detailed lean, with correct labels. If you mislabel your labels you can prove the moon is made of cheese in Lean.

Opus 5 says the disproof is wrong.

Complete bullshit. You don't know nontrivial from trivial groups LMFAO

This is entirely bullshit.

Nielsen is an accomplsihed physicist, and the person in here dunking on her is a hater who wanted to bang her a few years back. He had it set up to be his permanent post doc and hating on her forever is his love language. His nonsense here is entirely hearsay and no math is present.



Consider applying for YC's Fall 2026 batch! Applications are open till July 27.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: