Dan Larremore

My take, given the data: PNAS doesn't publish serious pure math anymore because math profs don't want to send their best work to PNAS. They prefer to send it elsewhere. But why did the emergent consensus change? No idea... Preprint here if interested: arxiv.org/abs/2603.00807

Consensus and fragmentation in academic publication preferences arxiv.org

arxiv.org/abs/1302.1886

Collective Motion of Moshers at Heavy Metal Concerts arxiv.org

After just a month of seeing loads of AI-generated slide decks, I yearn for a crystal clear, ugly, Beamer wall-of-bulletted-lists with equations and terrible graphics.

Slides at an academic conference in the age of AI… Claude has an incredibly obvious style, which is better than average but undifferentiated! needs bigger fonts. 🫣

AI Labs need to immediately recruit more Millennial and GenX parents to alignment. Years of XP being red-teamed by sentients w/ deception & willingness to mislead, for causes ranging from trivial to serious. Brushed your teeth, did you Astra? Well then you won't mind if I inspect this toothbrush...

If the speaker says “Oh dammit, how do I make the zoom controls disappear?” Then Zoom AI should automatically hide the goddamn floating controls bar that mars every hybrid meeting screen.