Remix.run Logo
▲ markasoftware 2 hours ago

I'm highly skeptical also, but OTOH extremely long human proofs have been a thing in the past, especially for important results like this.

▲lysace 2 hours ago | parent [-]

Published a number of hours after an AI-assisted proof has been published for a 150+ year old hypothesis? Doubt.

▲somesortofthing 2 hours ago | parent | next [-]

It's 940 pages long and presumably the process of producing it created many witnesses willing to vouch for its provenance as well as records of progress. Setting your math career on fire for a lie that obvious would be crazy.

▲zeraye 2 hours ago | parent | prev | next [-]

> The hope was to bring Alessandro to Stanford for a year to polish the writing and then announce it after running it by enough people. Now it seems best to put out the current draft.

We are not sure when this proof was drafted. It might have a few months or even more and it was rushed due to Alpöge announcement.

▲baruz an hour ago | parent | prev | next [-]

They have had another person reviewing it for months and the author has been incorporating his feedback, according to TFL. If anything I believe the AMS would give Masullo priority if both pan out.

▲igregoryca 2 hours ago | parent | prev | next [-]

It's a working draft, which they published hastily for obvious reasons.

▲adrian_b an hour ago | parent | prev | next [-]

It is said that others have been reviewing that human-written proof for many months.

Obviously, it was published immediately, despite the reviews not being finalized, when it became known that an AI-assisted proof was announced (but I understand that the complete AI proof has not been published yet).

▲dogecoinbase 2 hours ago | parent | prev [-]

The arrow goes the other way, most likely (similarly to what happened with Navier-Stokes).