LEAP panel report: Author predicted 35% on FrontierMath, actual 41%
dioscuri · x · 2026-08-29
The author shared the first accuracy report from the LEAP expert panel, reviewing their AI progress predictions:
- FrontierMath (end-2025): Predicted 35%, actual outcome 41%.
- GenAI-assisted US work hours: Predicted 5%, Fed reports 5.7%.
The author expressed satisfaction with the prediction accuracy.
More from AGI Musings
- Typing to AI is inefficient; voice interaction is the future — XFreeze · 2026-08-29
- Ramez Naam: 'Alignment' is incoherent, 'Instruction following' is the useful concept — jzl86 · 2026-08-29
- Carson Farmer: once intelligence gets cheap enough, we'll attach it to almost everything — carsonfarmer · 2026-08-29
- Opinion: Local Models Enable a New Class of Software with Embedded Intelligence — carsonfarmer · 2026-08-29
- Maybe AI Writing Is Bad Not From Missing Signals but Not Enough Smarts — jmugan · 2026-08-29
- London Futurists Podcast: Why optimists and doomers see different worlds — cccalum · 2026-08-29