With RH fallen, author says EA/AI safety is over-anchored on forecasters

JacquesThibs · x · 2026-11-21

After the reported Riemann Hypothesis proof, the quoted long post argues RH was the pre-named canary for dangerously advanced AI: the methods should generalize to anything definitively testable — including making AI better at making AI — likening it to a private firm producing a million tons of 15%-enriched uranium, not weapons-grade yet but a process line that could generalize to extinction-level threat. The author adds that forecasters still put RH at 2035 a week earlier, and says he long believed AI safety/EA is over-anchored on "forecasters" — a post he held back for being too spicy.

Original post →

More from AGI Musings

AGI Musings channel →