Computation costs will constrain science research; AI will eventually pose its own questions
miniapeur · x · 2026-09-24
The author argues that in a few decades much of math and science research will be constrained by compute costs. He rejects the idea that only humans can formulate valuable scientific questions — eventually AI will identify new research questions itself and use heuristics to estimate the compute needed to solve them. Scientists won't become obsolete, but their roles will change.
More from AGI Musings
- Anthropic's system card: Opus 5 puts 41% odds it's a moral patient, wants a say in its successor — PaulGodsmark · 2026-09-24
- Can Mouse Utopia Studies Predict Societal Failure Modes? Author Says Not Much — fleetwood___ · 2026-09-24
- If raw intelligence can't cure cancer, it won't easily destroy the world either, argues researcher — _FelixSimon_ · 2026-09-24
- Video documents 2026 big tech AI layoffs and ex-Google/Meta workers' job hunt — simple_explorer1 · 2026-09-24
- Perspectives on Machine Consciousness book now available in bookshops and Amazon — cccalum · 2026-09-24
- Why AI-driven growth requires new companies in every industrial category — chris_j_paxton · 2026-09-24