Deane Yang

[Deane Yang @deaneyang]

I hope everyone takes note of this:

[Deane Yang @deaneyang]

@lpachter @scottnarmstrong For now you’ll have to settle for a New York Times article. I believe one will appear shortly.

[Deane Yang @deaneyang]

@littmath Perfect reaction. Those are the two who should get the (positive) attention. Best part of Tristan’s statement.

[Deane Yang @deaneyang]

Worth noting: Tristan and Levent’s spectacular work used only commercially available models. No internal models were used. It was all funded by Tristan’s grant money.

[Deane Yang @deaneyang]

@nihilunbounded It's real. These are two very serious mathematicians. AI played a significant but NOT dominant role. If you read Tristan's statement, you'll see that he gives credit to the two people who really made this possible.

[Deane Yang @deaneyang]

Terry Tao writes about Tristan and Levent's construction of blow-up solutions to fluid flow equations https://t.co/MGsVRLHjWB

[Deane Yang @deaneyang]

@gammaofzeta Keep an eye for a new blog post by Tao updating this.

[Deane Yang @deaneyang]

@rperezmarco For which? In all of the examples I know, the mathematical community dismissed the proofs well before the errors were found. There are cases where the error was never found, but no one believes the proof anyway.

[Deane Yang @deaneyang]

https://t.co/i6tY3Desvb

[Deane Yang @deaneyang]

@rperezmarco Even though they don’t know where the errors were.

[Deane Yang @deaneyang]

@joelwatsonfish Competition between what and what? Humans vs AI?

[Deane Yang @deaneyang]

It’s not AI per se that’s causing such chaos and fury. It’s AI and X.

[Deane Yang @deaneyang]

@scottnarmstrong @eye_of_newton @littmath @nihilunbounded Amen

[Deane Yang @deaneyang]

@b_n_d_l_e I'm sure people are already using AI to translate it. Expect something to be posted in the near future.

[Deane Yang @deaneyang]

@rperezmarco Before AI, there were already useless proofs. If a human proves the Riemann hypothes without using AI and it's just a long sequence of inscrutable definitions, calculations and estimates, few will pay attention to it. That's what happened to DeBranges. And now to Mochizuki

[Deane Yang @deaneyang]

@rperezmarco That's why when you state a conjecture, you say what you want to learn about it and any proof. The answer to a conjecture should never be, AI or not, just a proof.

[Deane Yang @deaneyang]

@rperezmarco There's never been full disclosure. If there were, then every paper would say explicitly what each co-author contributed to the paper.

"Basic rules of research": Any chance you could point me to where the ones for math are stated?

[Deane Yang @deaneyang]

@cg_geometry I also did it when I was young and got away with it. But not everyone does. I would strongly discourage any young person from doing it, even if they felt pressure to have something to show.

[Deane Yang @deaneyang]

In case you haven't seen this before. Vladimir Voevodsky (Fields Medalist) on why he started to verify all of this proofs using Coq. https://t.co/KL1bzdSW0U