Agreed. The people need to know these cutting edge breakthroughs. @OpenAI just ask your clanker to write a popular math summary of your results and poast a Quanta article (btw automate @QuantaMagazine while you're at it, it's 100% automatizable and there's lots of revenue alpha there in making a popular science magazine that can pay off your compute debt) and communicate it directly with the commonfolk aka the Tax Payers. No need for mathematicians to be the middle men. Theorems, by the people, for the people, of the people.

Word on the street is that OpenAI gave a 25 page proof of Sogge's local smoothing conjecture which implies Bochner-Riesz conjecture which implies Fourier restriction conjecture which implies Kakeya maximal function conjecture which implies Kakeya (aka Besicovitch) conjecture, and it completely bypasses last two decades of painstaking progress in harmonic analysis just by doing elliptic PDEs and nukes the entire Kakeya pipeline from the orbit. I don't know any of this beyond superificial word of mouth, but I'm just putting this out there to say that maybe someone should check on Terry Tao.

Navier-Stokes SNAGGED Kakeya & restriction conjecture OBLITERATED 4D Yang-Mills existence SCOOPED TRIPLE KILL. ULTRA KILL. MONSTER KILL C C C C COMBO

Today I was trying to give a good, geometric proof of the Loomis-Whitney inequality. It states that for a measurable subset E in R^d, vol(E)^d-1 is bounded by the product of areas of the projections of E to the d coordinate hyperplanes. A very simple corollary of this inequality is the Euclidean isoperimetric inequality with non-sharp constants in every dimension, because area of the projection can be compared upto a factor of 2 with the area of the boundary of E.