We need to talk about AI and math. Most of the breakthroughs this year are counterexamples found by brute search, not theorems built from insight, and understanding that helps us grasp the power and limits of transformer models.
https://jonippolito.substack.com/p/ai-is-foraging-maths-meadows



![This quote:
Some users expected Maven to flag stale records or contradictions in the intelligence assembled for potential targets, though it is not clear why they thought the system would do that. The Minab [elementary school] site...came out as a recommended day-one target. Target-list work that once took hours was condensed into minutes....Staffing on civilian harm mitigation teams across the U.S. Department of Defense had also fallen by roughly 90% over recent years....The U.S. Central Command's group shrank from 10 people to one. No member of those teams reviewed the Minab site before the missiles were up in the air.
—2026 Bloomberg Palantir investigation](https://w3csocial.files.fedi.monster/cache/media_attachments/files/117/315/007/007/011/188/small/60cd5d31142cf9fc.png)




