Three of the eight OpenAI papers on the Hodge conjecture are withdrawn

Artisian 4 points 5 days ago

On the other hand, unique games still passes the sniff test so far. The model seems to have done a lot of work very quickly. This will change how research math gets done and funded, and we're seeing discussions of how it will change graduate studies in mathematics (perhaps the PhD defense itself will stop being ceremonial?)

save

path: 0 26224629, hotness: undefined, score: 4, children: 0

OwOarchist 4 points 5 days ago

So, it turns out it's much easier to hallucinate solutions to longstanding mathematical problems than to actually solve them. Who would have guessed?

save

path: 0 26225384, hotness: undefined, score: 4, children: 1

Artisian 2 points 4 days ago

Do note the other 300 papers. I agree we'll see more withdrawals, but the mathematicians I know seem to believe there is something in most of these papers.

save

path: 0 26225384 26241953, hotness: undefined, score: 2, children: 0

Treczoks 2 points 5 days ago

... are already redrawn. Lets see how many more won't survive scrutiny.

save

path: 0 26232682, hotness: undefined, score: 2, children: 0

lemonhead2 1 point 5 days ago

who found the error? did they find it themselves?

save

path: 0 26225915, hotness: undefined, score: 1, children: 1

Artisian 2 points 4 days ago

My guess is somebody on twitter. But I don't go there.