@ObhishekSaha What does it mean that you agree? Armstrong and Gallagher both have expertise directly relevant to Navier-Stokes, so they can have a meaningful difference of opinion. But you’re a number theorist
@bcubeddd@littmath Agree at least in as much I think it's very important to clearly distinguish issues with the publishing process from issues with the corpus of published work.
@littmath@bcubeddd I think this is obviously true in some fields (like AI itself) but I've always had the view that the math literature is pretty high quality, publications in the 'elite journals' are actually (usually) rather good, etc. In what sense do you think it was broken?
@deanwball So I guess I don’t know what you meant by “I’ve never felt history repeat itself more explicitly.” You don’t think Beethoven opus 12 are very good and you agree with both criticisms?
@nihilunbounded I’ve had the impression that he has some funny perspective that math is a subfield of biology. No idea about his current views on LLMs though
@LH The importance of this problem is to do with understanding mathematical techniques of analyzing PDE. Anything about the “actual physics of the universe” would be a (great) bonus
@Quasilocal While staying agnostic on any particular instance, it seems kind of inarguable to me that it happens... though still overstated in some quarters
@nihilunbounded Hmm. Perhaps am misreading your initial post and/or Poincaré's proof of the uniformization theorem but I believe the key part of the latter was his cutting edge work in potential theory and the Dirichlet problem
@Quasilocal Probably you've already seen, but this new paper suggests the possible relevance here of training on unpublished preprints. See the AI statement on the last page https://t.co/bEB1B9HPEa