Mathematician @ElliotGlazer explains why the 108-page Alpöge paper shouldn't be 108 pages: "People working through it say it should not be this long. The key ideas are actually all in the first few pages. Claude is being very pedantic about the wrong things." "Literal mathematical mistakes are much rarer. Expositional faux pas are very common. Not putting enough details in the hard parts, putting too many details in irrelevant parts. AI-generated papers are plagued by this, and this one…