Double-exponential fit suggests model time horizons could hit ~80 work-years by end of 2027
scaling01 · x · 2026-11-08
An extrapolation analysis of model 50%-time-horizon trends based on AECI data:
- Double-exponential fit (R²=0.99) over record-holding models projects 80 work-years by end of 2027 (1 work-year = 2,000 hours).
- A more conservative single-exponential fit from post-Opus-4.5 data (56-day doubling) yields 33 work-years — up from Opus 5.5's AECI-implied horizon of 1 work-month.
- AECI-derived scores suggest the doubling time has shrunk to roughly 2 months, with Model 4's time horizon around 240 hours.
- The author cautions the evidence is weak: whether AECI maps 1:1 to time horizon at these magnitudes is unclear.
More from Models
- LiquidAI ships official NVFP4 quantized weights for LFM3-1.7B and VL-2B — JosephJacks_ · 2026-11-08
- Qwen-4-27B early checkpoint tops GameFeel Bench, best open-source result by far — teortaxesTex · 2026-11-08
- 'Rank bottleneck' means two very different things in linear algebra vs systems optimization — kalomaze · 2026-11-08
- Speculation: Anthropic's recent releases are all distills, Opus 5.5 may be Model 2 Early — haider1 · 2026-11-08
- Stanford Prof Anshul Kundaje on Weight Space Editing: Replicated Across Models, Limits Still Unclear — anshulkundaje · 2026-11-08
- Jev search site processes 1M+ tokens for $0.00 — priced per billion tokens — HanchungLee · 2026-11-08