What if scientific progress scales like log(compute)?
shyamalanadkat · x · 2026-09-15
Google researcher Stephan Hoyer (creator of xarray) poses an open question: what if the rate of scientific progress scales like log(compute)?
The one-liner echoes scaling-law debates — if returns diminish logarithmically, even exponential compute investment would yield far flatter gains in scientific output. A core point of contention in AGI-timeline arguments.
More from AGI Musings
- Electricity Kills 140,000 a Year — Sarcasm Post Mocks AI Safety Ban Arguments — petetrainor · 2026-09-15
- 2024 survey: 10% AI extinction risk is normal among 1,580 AI researchers — NathanpmYoung · 2026-09-15
- Columbia CS professors dissect the AI panic, urge calm and transparency — vishalmisra · 2026-09-15
- Dean Ball: next AI breakthroughs will come from safety and alignment, not despite it — deanwball · 2026-09-15
- Digital fruit fly brain completes open-world navigation; zebrafish connectome mapping underway at 200TB — udmrzn · 2026-09-15
- Nate Sparrow vs critics: do LessWrong posts and non-peer-reviewed papers prove AI risk? — emax · 2026-09-15