This is a special post for quick takes by leogao. Only they can create top-level comments. Comments here also appear on the Quick Takes page and All Posts page.
the new openai planar unit distance result kills my last remaining doubts about AI being a huge multiplier on research productivity in the near term future. i was not expecting this to happen so soon; i would have guessed probably another year before we got a result like this.
It seems like this is still a case of solving an easily-verifiable problem, so I'm not sure how much that should generalize to solving less-verifiable problems. It feels like verifiable problems were on the path to being completed solved as soon as the Ralph Loop came out.
What about it in particular convinced you that e.g. the previous big result didn't?
13leogao
i get the impression that the previous problems were mostly just neglected, or otherwise were less impressive than they seemed. whereas afaict mathematicians agree the new result is on a real well-known problem and genuinely surprising and novel.