Researcher calls for making recursively self-improving AI illegal
harris_edouard · x · 2026-10-08
Echoing Esben Kran (@EsbenKC): commenting on any development in AI now feels hard, but the one thing he knows is that we must make it illegal to create recursively self-improving machine intelligence — a hard-line AI safety stance on frontier capability growth.
More from AGI Musings
- Megaconstellation math: 1.4M satellites would burn 520K tons per year in atmosphere — bradneuberg · 2026-10-08
- Zak Kohane: AI will hit surgical intervention sooner than many medical subspecialties — zakkohane · 2026-10-08
- VectorInst asks: what physics rule violation still shows up in generative video? — VectorInst · 2026-10-08
- Economists Use Firms' Historic University Hiring Patterns as an Instrument for AI Adoption — TaniaBabina · 2026-10-08
- AI delegates as smart voicemail: imagining an audio-native UX along async/sync gradients — curious_vii · 2026-10-08
- A Lee Sedol moment for math: AI proofs will be free for all, raging at OpenAI is pointless — RexDouglass · 2026-10-08