MC

[MC @MCktm5]

@normonics decentralizing? how are LLM's, which are under the complete control of very few companies and require masssive data centers, decantralizing anything? LLM's are centralizing all wealth and power into the hands of a few tech people.

[MC @MCktm5]

@__paleologo @W98AB Yes, but their predecessors cite them in papers, while LLM's don't.

[MC @MCktm5]

@cloneofsimo He's talking about Pure Math Research, not the curing of diseases or resolving global problems. How about OpenAI spends Billions of dollars on trying to do that, instead of solving math papers very quickly before mathematicians finish their own.

[MC @MCktm5]

@haider1 Who is this "we"? What posts are you talking about that treat Mathematcians as authorities on future LLM capabilities? The mathematicians I know, usually have the opinion that predicting the future is too hard to say anything about.

[MC @MCktm5]

@nealasher Yeah, a leading expert, talking about their own field is a moron.

[MC @MCktm5]

@aryehazan Why are people having such difficulty having some trust and care for the Mathematical Community. This is the community that created the framework for LLM's to even work in the first place. It's the community that built the theory, that serves as the backbone for the Sciences.

[MC @MCktm5]

@proofchecker @shallit43 They can give the talk, but publishing is for novel results.

[MC @MCktm5]

@sytelus @shallit43 Yes, but the talk was merely announced, presumably the paper would have been up by the talk.

It's not unreasable to schedule talks, ( need lots of organizating) before the paper is fullly polished.

[MC @MCktm5]

@AntoineTilloy That's how math has been since humans started counting with their fingers. Sorry, that the education system has given you a different impression of what math is.

[MC @MCktm5]

@mouse_math yes. and semi-literate is very possibly an overestimate.

[MC @MCktm5]

@ericxing Mathematicians aren't blaming LLM's, they are blaming the enormously greedy,profit motivated, LLM Companies.

[MC @MCktm5]

@JTBeers @kareem_carr Yes, but the meaning is, I applied X Linear Algebra to produce the symbol Y. I don't see any reason to believe that the meaning of Y, and the Linear Algebra that produced Y, to have any coherent relationship. Y was produced because in similar cases Y was likely the right answer

[MC @MCktm5]

@GregoryConti19 What are you talking about? What is a better use of ones time then developing ones own abilities and understanding?

[MC @MCktm5]

@ramez @dbinthed maybe, but they don't have the right to claim ideas that aren't theirs. It's still academic fraud. Are you even capable of understanding if they stole an idea? Would you recognize if a mathematical construnction came from one of the papers?

[MC @MCktm5]

@venturetwins @ben__bradley I think you are very confused about the complaints. The problem isn't LLM's solving problems, it's that the companies are using them rampatently, in an unethical way that could cause great harm.

[MC @MCktm5]

@MatthewTanous What are you talking about? Most mathematicians post their work for free public acccess on the arxxiv server for everyone to see, and present their research ideas in talks that are recorded and put on youtube.

[MC @MCktm5]

@JTBeers @kareem_carr LLM's don't use human techniques to solve problems, they apply an algorithm to predict the next symbol to generate. Them telling us how they do what they do is worthless.

[MC @MCktm5]

@Math3matica @brandonjcarl And you base this prognosis on what?

[MC @MCktm5]

@scottstts Nobody grows, when someone else
solves the problem for them.

[MC @MCktm5]

@scottstts Who in the "real world" was waiting to see if NV can be made to blow up in finite time? We already knew that NV doesn't apply when forces exceed speed of sound. People grow when they solve problems, that's the main point of the Millenium Problems, to help the math community grow