Acemoglu's Fifth Question on AI: Can We Avoid Centralized Control of Intelligence?
joshgans · x · 2026-10-06
In the fifth of his series of questions on AI, Daron Acemoglu asks whether we should accept the centralizing trajectory of AI — ever-larger models trained on much of the world's knowledge, run on concentrated compute by a few companies — or whether AI's promise can be realized with data, capabilities, and control far more decentralized across domains and people.
Economist Josh Gans responds that decentralization and human agency are obviously appealing in principle, but murky in practice. He notes that many technologies concentrate power through scale economies and network effects, and AI companies appear to be heading the same way, despite early computer pioneers like Lee Felsenstein believing digital tech would be powerfully decentralizing.
More from AGI Musings
- YOLO-mode LLMs create an absurd capability gap against IT-deficient counterparties — curious_vii · 2026-10-06
- Should AI Triage Mental Health Crisis Messages? The Case for Human Review — yi111 · 2026-10-06
- Neuroscientist argues experimental neuroscience can't crack intelligence, but aids disease therapies — aran_nayebi · 2026-10-06
- Bryan Johnson: the global era of Don't Die begins in 2030 — Zachly · 2026-10-06
- Who believes in fast foom but near-zero doom probability, with deep technical arguments? — burny_tech · 2026-10-06
- AI narrows the visible beginner-expert gap, but taste and judgment now separate people — r0ck3t23 · 2026-10-06