Security Expert Maps the Agent Safety Stack: Security, Alignment, Policy, and Why CoT Won't Save Us
joshua_saxe · x · 2026-09-04
Security expert Joshua Saxe synthesizes the babel of AI safety talk into six points: security (sandboxing) is the only source of hard guarantees; alignment can never be deterministic; policy must pace agent autonomy; chain-of-thought is a low-signal, 'limited time offer' signal; it will eventually be replaced by reading intentions from latent spaces Minority-Report style; and the community's rush to --dangerously-skip-permissions is eroding the one layer that actually works.
Related event: Security veterans untangle AI safety: security vs alignment vs policy(9 posts)→
More from AGI Musings
- AI agents are erasing work-life boundaries: managing agents on commutes, vacations and at 3am — bendee983 · 2026-10-21
- Stanford's Kundaje: genomic AI models must test cell-type specificity to expose real weaknesses — anshulkundaje · 2026-10-21
- a16z: Superhuman AI hardware design is the easy part — manufacturing is the hard one — a16z · 2026-10-21
- a16z's Casado says AI can't do a VC's job — but it can do everything around it — alvelda · 2026-10-21
- Kent Beck: Let the Genie Carry the Toil, But Keep the Struggle — josh_wills · 2026-10-21
- Nearly $3B mapped: the money behind the AI existential risk ecosystem — KeyNaive2982 · 2026-10-21