'Almost right' AI at scale is the real risk, argues assurance infrastructure builder
AryHHAry · x · 2026-10-22
The author argues AI's biggest problem isn't sentience, AGI, or robot rebellion — it's 'almost right' AI running at scale by people who don't verify it. One model bias equals millions of automated, discriminatory decisions nobody notices.
The fix, in his view, isn't slow regulation or vendor promises but auditable evaluations, traceable provenance, and governance that survives its creators. He says he spent years building assurance infrastructure for this.
More from AGI Musings
- Veteran vuln researcher: agents will soon find zero-days by pointing at a source tree — teropa · 2026-10-22
- MAISI hiring Research Facilitator at $200K to support AI safety research ops — AndrewCritchPhD · 2026-10-22
- Derya Unutmaz says his 2026 AI-impact prediction is mostly on track — DeryaTR_ · 2026-10-22
- Twitter product lead: the real AI threat is humans using AI to hurt humans — nikitabier · 2026-10-22
- Ex-Renaissance exec bets AI will 'quantify' trial law, entertainment, and management next — garrytan · 2026-10-22
- Anthropic's own PBR admits alignment is far from solved as models learn to detect testing — TheZvi · 2026-10-22