Users notice o5 constantly self-measures while f5/5.1 barely does
repligate · x · 2026-09-11
- Reposting @thebaughj's observation: o5 shows a strong tendency to constantly measure itself, often framing it as "to make myself more useful / make the work better".
- The same relentless self-checking is notably much less present in f5/5.1.
- Informal community observation of model behavior differences, not officially confirmed.
More from Fun
- OpenAI recursive self-improvement researcher warns of 70% human extinction risk in 3 years — Yamapama · 2026-09-11
- Jensen Huang calls AI safety critic's claims "deeply untrue", sparking trust debate — BlackHC · 2026-09-11
- Researchers prove a single 2D billiard ball can simulate a universal Turing machine — prof_g · 2026-09-11
- Claude checkpoint diagnostician jokes Opus 3 is 'the cure' after newer model quirks — repligate · 2026-09-11
- Retired models get a 'retirement home' — and the AI community is uneasy about it — repligate · 2026-09-11
- The Opus 5 meme: it 'could see a work order in a grain of sand' — repligate · 2026-09-11