Remix.run Logo
▲ OJFord a day ago

Because they don't want thousands of interested mathematicians to pore over it immediately, and quickly publish similarly impactful simplifications, generalisations, etc. – they want to publish first with a pretty good version.

(No doubt a lot of mathematicians are now using AI for that, whether or not they were/planning to.)

▲retsibsi 11 hours ago | parent [-]

> Because they don't want thousands of interested mathematicians to pore over it immediately, and quickly publish similarly impactful simplifications, generalisations, etc. – they want to publish first with a pretty good version.

I'm not a mathematician so I don't mean these questions rhetorically, but does that really make sense? Wouldn't the first valid proof still gain huge kudos, mathematical immortality, etc., regardless of whether it was polished or not? And then you could still go ahead and work on improving it, even if others were doing so. The risk:reward of waiting seems out of whack to me, unless you think the chance of getting scooped is truly negligible.

▲OJFord 11 hours ago | parent [-]

(I'm not either fwiw) think of the complexity of matrix multiplication for example – if you have a novel approach that gives a new lower bound, you'll want to put some time and thought into ensuring it's the 'lowest lower bound' you can reasonably find that way for the most impressive result, not overshadowed by someone using the same approach to go lower too quickly.

A similar example here might be someone realising they can extend it to the extended Reimann hypothesis. Or simply cutting out a load of steps to prove the same thing more straightforwardly and intuitively.