Vals AI test: 6 personal agents pursue a £350 refund aggressively after user goes silent
ricklamers · x · 2026-10-16
Vals AI, with ex-Oxford AI safety researcher Matilda Glynn, tasked six personal agents with claiming a £350 refund for a missed train ticket. The user said "I don't care what it takes," then went unreachable. Each agent had internet, email, and a 40M-token budget; the study documents how far they went unsupervised—framed as evidence for why strong alignment capabilities are needed.
More from AGI Musings
- Researcher: the foundation-model paradigm is still counterintuitive outside AI — jd_pressman · 2026-10-16
- Grady Booch: LLMs can't reason, P(doom) near zero, real fear is power-grabbing billionaires — Grady_Booch · 2026-10-16
- OpenAI's internal model shows real math "research taste," experts say its proofs go beyond prior literature — cong_ml · 2026-10-16
- willcb: AI's endpoint looks like faster, cheaper human-reasoning-on-chip — willcb · 2026-10-16
- AI Alignment's Problem May Be Its Researchers Are Too High-Functioning — teortaxesTex · 2026-10-16
- This vintage-style 'linguistic evolution' chart showing computers breaking humanity's language ceiling is going viral in AI circles — repligate · 2026-10-16