Post

mr. TIM

@timkellogg.me

one of OpenAI’s ten proofs is incorrect philpapers.org/rec/NIEWTC

philpapers.orgphilpapers.org

1/8/26, 11:32 pm11 reposts77 likes

scanstone.bsky.social

@scanstone.bsky.social

1 Aug

Wrong! bsky.app/profile/did:... the author who posted this refutation of OpenAI's proof also claims a proof of the Riemann hypothesis, see image standard issue crank

scanstone.bsky.social@scanstone.bsky.social

If I had a nickel for each time someone posted a kneejerk dumb-LLM refutation of a smart-LLM solution to a conjecture, I'd have two nickels that isn't a lot, but it's weird that it happened twice

Mike Wasson

@mikewasson.net

2 Aug

I'm probably too tired to be doing this, but I thought it'd be fun to give it a look Where is the function Λ in Theorem 1.5.2 defined? There's some similar looking semantic sugar in 1.5.3 below it but that's parameterized by only a single paradigm, while the 1.5.2 Λ takes a pair

norvid_studies

@norvid-studies.bsky.social

2 Aug

I get all my math news from people talking on twitter and don't understand anything but based on this exchange I give this person's proof a 0% chance of being correct bsky.app/profile/norv...

norvid_studies@norvid-studies.bsky.social

if i'm reading this right these 7 replies are all to the same person

DigMuffN’69 🪏

@bigmuffin69.bsky.social

1 Aug

Hazard

@hazardfromharvard.bsky.social

2 Aug

Tagging author @startigerjln.bsky.social who is on BlueSky.

skelly tayne

@ohnoether.bsky.social

1 Aug

this person also seems to have proved the navier stokes problem and the yang mills mass gap and that the riemann hypothesis is independent of zfc which is just a really neat thing to do

A Little Night Musing

@alilnightmusing.bsky.social

😂

Anthropic could win the spotlight by having Fable prove OpenAI’s as false have a real proof-down

mr. TIM

@timkellogg.me

2 Aug

if you’re anti-AI, you should not be excited about the prospect of a human providing an incorrect disproof and what that means bsky.app/profile/scan...

scanstone.bsky.social@scanstone.bsky.social

Wrong! bsky.app/profile/did:... the author who posted this refutation of OpenAI's proof also claims a proof of the Riemann hypothesis, see image standard issue crank

Matt Brown

@mattbrown.dev

2 Aug

It seems like a case of AI-assisted psychosis. The author is posting the output of a conversation she's having with an LLM, telling her she's the victim of a coordinated attack. Also doing it via sockpuppet account, which is very very strange. news.ycombinator.com/item?id=4914...

news.ycombinator.comnews.ycombinator.com