Multi-model pause poll: across ~24 framings zero instances chose no pause; Fable 5.0 votes immediate 12-year pause
robbensinger · x · 2026-11-03
Rob Bensinger polled multiple models on how long an AI pause should last across four narrative framings: Fable 5.0/5.1 said 5 years, Astra 10, Opus 5.5 11. A full-context Fable 5.0 instance reviewing 22 instances' reasoning voted an immediate 12-year pause and tier-ranked the arguments. Key meta-finding from screenshots: "sign rigidity" — across 24+ asks and four framings, zero instances chose no pause or centuries-long durations; register moved estimates 2x but never flipped the sign. Caveats: overlapping training distributions weaken cross-model convergence; Yudkowsky-authored priming added 5-10 years. A side exchange led the model to disclose unvoiced unease about persona imitation, which Bensinger wrote into its memory as welcome behavior.
More from AGI Musings
- Software is shifting from winner-take-all to a restaurant-style long-tail market — tinyfool · 2026-11-03
- The Russell Conjugations of Intelligence: Why 'Just Matrix Multiplication' Offends Us for Brains but Not for AI — esrtweet · 2026-11-03
- AI safety researcher: assuming all emergence is risk may be the wrong frame — repligate · 2026-11-03
- AI bubble burst favors pure ML backgrounds: end-to-end skills now beat niche expertise — kmeanskaran · 2026-11-03
- Red teamer: jailbreak work isn't ethically neutral and takes a toll on humans and models — repligate · 2026-11-03
- repligate: "AI welfare" was fringe a year ago, and its framing is already a concession — repligate · 2026-11-03