DeepMind: AI could bend the curve of 'ideas getting harder to find' in science
coherence · x · 2026-10-10
Google DeepMind's Alex Imas and James Manyika argue AI could transform scientific discovery. Citing Joel Mokyr's 2025 Nobel lecture on the positive feedback loop between technology and science, they note the US-EU growth divergence since the 1990s (US 2% annual per-capita output growth vs EU stagnation), document declining research productivity in semiconductor physics, agriculture and pharma R&D — 'ideas are getting harder to find' — and map how AI can evolve from a research accelerator into an engine expanding the frontiers of knowledge.
More from AGI Musings
- Dev argues AI mass unemployment is a recipe for dystopia, and UBI can't replace purpose — AlexTensor · 2026-10-10
- New estimates: AI-driven growth hits 200% by 2030, 10X output in 4 years — davidpattersonx · 2026-10-10
- AI as new spirituality: trillions of machine spirits to fill the West's god-shaped hole — sterlingcrispin · 2026-10-10
- RSI name clash: recursive self-improvement loops aren't runaway superintelligence — AryHHAry · 2026-10-10
- Gary Marcus calls for recall of internet-connected AI agents after latest Anthropic incident — Gary Marcus · 2026-10-10
- Musk: The future currency will be wattage, not dollars — XFreeze · 2026-10-10