The Strip Mining of Mathematical Proofs (with Hugo Duminil-Copin) - Numberphile Podcast

Today I'm speaking with a French mathematician Hugo Dumman Copan, the 2022 Fields Medalist. I'll be talking to him about the very recent release by Open AI of hundreds of mathematical proofs. It's a truly seismic development in the math world. Hugo is someone I particularly wanted to speak with because at least 16 of the proofs that

were published were really things directly working on. Put simply, problems in statistical physics, depending on your viewpoint, he's been one of the biggest victims or beneficiaries of this math proof tsunami. But Hugo also made headlines and cpped some flak on social media when he was quoted saying that he'd cried when the results were published. Proper

out loud crying by all accounts. I wanted to find out more about that, too. [music] Let me ask you then about this day in October that every mathematician is talking about when all of these proofs were released by Open AI. Has it got a name? What do you What are you all calling it? I've seen like the slop drop

and the math apocalypse. What do you call this day? >> Yeah, I don't have a special name for it because I mean they they come with a connotation, right? So probably it's better not to give it a name, but yeah, it was a day where they release all these papers. So it was clearly

uh a moment for mathematics. Yeah. Can you give me some idea how big a day in mathematics you think this will be historically like this this feels like a seismic day like like it will be a day that's looked back on as a a turning point for better or worse. >> Yeah, I mean exactly. I think it's

difficult to know where it's going to go, but clearly automatization came into the into our world, right? I mean, now like the computers are more than just helping. They are in some sense leading and we have to uh to catch up and to be sure that it's used in a in a good way for mathematical