Six months after 'too dangerous to release,' open weights wreck only mathematicians
ziv_ravid · x · 2026-10-07
A satirical post: six months after a frontier capability was deemed too dangerous to release for cybersecurity reasons, models at that level are now open-weight (GLM 5.3) — and the damage report reads: power grids taken down 0, banks collapsed 0, mathematicians made redundant via 722 AI papers and counting. "We red-teamed the wrong department." A much-shared jab at safety conservatism versus actual AI risk.
More from Fun
- Matt Slotnick mocks the endless "Google should buy Instinct" hot takes — matt_slotnick · 2026-10-07
- "Can I pick your brain?" is just human distillation — signulll · 2026-10-07
- Same prompt, different results: nanobanana image generation compared day over day — leslysandra · 2026-10-07
- 'Every job in my field is temporary': tradesman rebuts reporter on data center jobs — csuwildcat · 2026-10-07
- Zombie apocalypse short made with Flovaai shows off AI video scene consistency — thetripathi58 · 2026-10-07
- Instagram Reels clips of old films in AI-looking 60fps raise authenticity concerns — Svenray · 2026-10-07