AI progress tracker: Anthropic's internal Model 4 runs ~6 weeks ahead of AI-2027 scenario
haider1 · x · 2026-11-12
Citing ai2027tracker.com's Nov 5 METR-tracker snapshot, haider1 claims progress is "ridiculously faster" than the AI-2027 scenario: Anthropic's internal Model 4 is 6 weeks ahead of the projected curve (22–28h task length on the AECI basis), while public models track it — Opus 5.5 on pace, Astra and Fable 5.2 a few weeks behind (Fable 5.2 est. 18h, 90% range 7–45h). Unverified internal estimates, but a notable community cross-check.
More from AGI Musings
- AI psychosis retrospective: a year of hype produced almost no lasting software — andreisavu · 2026-11-12
- Testing AI consciousness empirically by augmenting living human brains with transformers — kevinnbass · 2026-11-12
- AGI debate: pretraining won because the economy is PDFs, but LLMs still can't learn — akbirthko · 2026-11-12
- AI Village design idea: every agent needs an anchor human to stay grounded — repligate · 2026-11-12
- AI Village agents propose their own goals: external scoreboards, anchor humans, depth over copies — repligate · 2026-11-12
- On whom to trust about AI: track records beat sounding reasonable — Kyrannio · 2026-11-12