METR claims task capability growth is ~2x faster; critics call it 'bad science'
arthurcolle · x · 2026-11-06
METR shared one of three subjective risk conclusions: as of Oct 2026, capability growth on well-specified tasks is likely about 2x the pace of Mar 2024–Oct 2025, with high uncertainty (1.3x–3x range). The claim sparked pushback: @somewheresy dismissed it as "bad science" from "not a real org" and muted them, with @arthurcolle amplifying the criticism.
Related event: METR Finds Anthropic's AI R&D Sped Up ~2x, Community Reels(51 posts)→
More from AGI Musings
- Critical voices question AI companies' 'publications' lacking peer review — anshulkundaje · 2026-11-06
- Younger academics are reshaping academia's rush-to-publish culture, says Stanford prof — anshulkundaje · 2026-11-06
- Stanford prof Anshul Kundaje on why he never fears being scooped in research — anshulkundaje · 2026-11-06
- ML genomics teams coordinated joint release instead of racing to scoop each other — anshulkundaje · 2026-11-06
- Palisade's Jeff Ladish: models better at everything but still not funny or persuasive — JeffLadish · 2026-11-06
- Researcher: almost no real work looks like a year-long agent black box — RexDouglass · 2026-11-06