Stuart Russell: The real problem is AI technology is intrinsically unsafe
ronbodkin · x · 2026-09-22
Stuart Russell, UC Berkeley AI professor and president of IASEAI, argues that the core problem today is that the technology is intrinsically unsafe: we are scaling AI systems we still don't fully understand. Making AI more powerful, he says, is not the same as making it controllable — safety requires understanding the systems we build.
More from AGI Musings
- Commenters spar over China's open-model lead and whether the US would nationalize Nvidia — AIFlow_ML · 2026-09-22
- Schmidhuber: I published the first concrete RSI algorithms back in 1987 — SchmidhuberAI · 2026-09-22
- French essay: AI's gravest peril isn't doing our tasks but sparing us the inner transformation they demand — anselm · 2026-09-22
- 10k agents compress a century of work into 88 hours: multi-agent swarms' capability and risks are underestimated — mattbeane · 2026-09-22
- MIRI president Nate Soares says he can no longer rule out an intelligence explosion within six months — victor_explore · 2026-09-22
- Could an agent swarm replicate 95% of YC startups? Capability isn't commitment — abhiadesai · 2026-09-22