Marc Andreessen: both AI utopians and doomers will be disappointed — nuclear is the precedent
sudoraohacker · x · 2026-11-17
Marc Andreessen restates his thesis: both AI utopians and doomers will be disappointed — specifically by the scope and magnitude of AI's real-world and economic impact.
- The world is complicated and messy; an obvious deployment path for a new technology doesn't mean it actually happens;
- The historical precedent is civilian nuclear power, once seen as the panacea for US and global energy growth, frozen in its tracks before delivering;
- The implication: AI's fate may hinge on real-world deployment friction, not the technology alone.
More from AGI Musings
- Panagram slop detection could collapse the anti-AI argument, argues MLStreetTalk — burny_tech · 2026-11-17
- From Artificial Intelligence to Artificial Intimacy: Viral Take on AI Companions — dhruv2038 · 2026-11-17
- Yann LeCun publishes his United Nations Security Council address as receipts — ylecun · 2026-11-17
- MIRI launches "Do Not Accept" campaign against AI companies deciding humanity's future — zetalyrae · 2026-11-17
- Jane Street's yminsky: agents are bad at software design — but does design still matter? — charliermarsh · 2026-11-17
- WSJ credits Gemini with diagnosing what dozens of doctors missed — Redditors question the framing — Aaasteve · 2026-11-17