Manheim: real AI research horizons may be 100x shorter than METR's, at same growth rate
davidmanheim · x · 2026-11-11
Researcher David Manheim argues, based on preliminary real-world data, that AI time horizons on actual research tasks are 1-2 orders of magnitude shorter than assumed by AI 2027, METR, and ECI-style forecasts—and that ECI doesn't map cleanly to exponential gains in real time horizon or researcher productivity. He calls the claims he's responding to slightly overstated but roughly correct, stressing the key point: real research horizons could be 100x shorter than METR horizons today while growing at the same exponential rate—so forecast correctness hinges on the calibration baseline, not the growth trend.
More from AGI Musings
- teortaxesTex: RL may give coding models 'full fluency' with no superhuman headroom — teortaxesTex · 2026-11-11
- Frontier AI Still Lacks 'Heresy': Breakthroughs Come from Defying Consensus — JeremyNguyenPhD · 2026-11-11
- Dev argues optimism is the only trait that matters as intelligence gets commoditized — zack_overflow · 2026-11-11
- AI risk community splits over 'intelligence explosion next year' claim as credibility war erupts — connoraxiotes · 2026-11-11
- AI lets robotics research thrive in labs with no robotics experts — 4310sy · 2026-11-11
- Dictating notes to an AI agent feels as mundane as putting on shoes — _FelixSimon_ · 2026-11-11