>The researchers place their work in the broader context of what they call vibe-proving, an approach in which language models are used to explore and structure high-level theoretical reasoning. https://phys.org/news/2026-03-chatgpt-mathematical-proofs.html
Eat shit and die, loser. That is all I have to say.>t. arithmetic geometer
>>16932780Nice proof, fag
>>16932779>original mathematical proofsI would hope so, since math, like coding, is entirely algorithmic and reducible, especially with an LLM, to logic-gate representation.
>>16932784>proofs are algorithmic You...I like you....
>>16932787the proofs aren't algorithmic, the logic is, hence an LLM's ability to derive different or novel proofs by optimizing the logic.It's like chess, we shouldn't be surprised that computers long ago exceeded the strength of the best grandmaster.
>proofsfaggot
>>16932780Why are you so angry?
>>16932797>algorithms aren't algorithmic, algorithms for producing the algorithms areThis just blew my mind
>>16932779How can I use it to say game the stock market?
>>16932779lmao bye miggers
>>16932797Little do you know that the best chess bots got beaten bunch of times by a matures because they were trained too much on beating grandmasters and other AIs.>>16932989Stock market can’t be gamed, it’s 99% about luck or insider trading. Random roll of dice often beats the best investors. The only reliable strat is spread your investments if there is bull market and buy metals/bonds when there is bear market.
>>16932779may I see these original proofs?
>>16932989The sock market is not a mathematical system just because it uses numbers
>>16932779>original mathematical proofThis is a meaningless criterion. Token guessers generate "original proofs" the same way they generate anything else "original": trivial alterations and recombinations of 500 TB of human data. They still reliably fail at any out-of-distribution problem, even such that are accessible to elementary school children. It would be vastly more useful to have a semantic search engine that can reliably pull up potentially relevant mathematical results given a problem and let actual intelligence do the reasoning, because the only advantage token guessers have over mathematicians is superhuman access to prior results.
>>16933566Nigger. Read the Man who solved the market
