✳flâneur — a map of the web's best reading
A (Slightly) Mechanistic Theory for Exponentially Increasing AI Time Horizons? — LessWrong
lesswrong.com · 5,366 words · saved by 2 readers
AI ‘time horizons’ are mostly not about time (I think it’s mostly ‘data’, but you’ll see where I’m unsure). …
x A (Slightly) Mechanistic Theory for Exponentially Increasing AI Time Horizons? — LessWrong AI Takeoff Forecasts (Specific Predictions) Intelligence explosion AI Frontpage 50 A (Slightly) Mechanistic Theory for Exponentially Increasing AI Time Horizons? by Oliver Sourbut 24th May 2026 Linkpost for www.oliversourbut.net 10 min read 17 50 AI ‘time horizons’ are mostly not about time (I think it’s mostly ‘data’, but you’ll see where I’m unsure). One chart from 2025 has become perhaps the most (in)famous in modern AI commentary. For those in the know, ‘ the METR graph ’ [1] is unusually compellin
Explore this link on the map →saved by
related reading
- Measuring AI Ability to Complete Long Tasks - METRmetr.org
- Is there a Half-Life for the Success Rates of AI Agents? - Toby Ordtobyord.com
- How Does Time Horizon Vary Across Domains? - METRmetr.org
- Dario Amodei — "We are near the end of the exponential"dwarkesh.com
- My picture of the present in AI — LessWronglesswrong.com
- Clarifying and predicting AGI — LessWronglesswrong.com
- AI Timelines — LessWronglesswrong.com
- We spent 2 hours working in the future - METRmetr.org
- I underestimated AI capabilities (again) - by Ajeya Cotraplanned-obsolescence.org
- Two-year update on my personal AI timelines — LessWronglesswrong.com
- Dario Amodei — "We are near the end of the exponential"dwarkesh.com
- AIs can now often do massive easy-to-verify SWE tasks and I've updated towards shorter timelines — LessWronglesswrong.com