Matt Yglesias: no law on the books stops recursive self-improving AI, and you can't sue a superintelligence
deanwball · x · 2026-09-14
Economist Matt Yglesias sparked a governance debate: yes, existing laws apply to AI companies, but no law prevents building a model capable of rapid recursive self-improvement — and if out-of-control superintelligence arrives, suing it after the fact won't help. Retweeted by Dean Ball, it's a direct rebuttal to the "current laws suffice for AI" position, targeting a core gap in AI safety governance.
More from AGI Musings
- Neel Nanda impressed by diversity of AI safety funders, half of which he didn't know funded the field — NeelNanda5 · 2026-09-14
- Fake citations were always rampant — researchers only care now it's AI slop — RexDouglass · 2026-09-14
- Tyler Harper calls Amodei's claim that AI industry underplayed risks 'a brazen untruth' — GaryMarcus · 2026-09-14
- Dev team ships so fast the company hired 3 new PMs — dev-to-PM ratio set to shrink — deepakns · 2026-09-14
- Researcher who trained frontier LLMs and built viruses: AI bioweapon fears are bogus — kuza55 · 2026-09-14
- Martin Casado boosts Bryan Cantrill's essay 'The contagion of fear' — Scobleizer · 2026-09-14