AI safety researcher: catastrophic and existential AI risks demand different policies
davidmanheim · x · 2026-10-27
In a thread with Vox's Bryan Walsh, AI safety researcher David Manheim argues opposition to AI x-risk often stems from the concept itself—Parfit-style reasoning aside, once you're in catastrophe territory the practical actions look similar. But key divergences remain: for AI, many claim catastrophes can be prevented with human oversight and domain restrictions, so recursive self-improvement is deemed fine. He separates biosecurity from AI risk, citing his CSER talk and new surveillance paper.
More from Safety
- Matthew Green recalls being brushed off by OpenSSL weeks before Heartbleed hit — matthew_d_green · 2026-10-27
- Tailscale confirms privilege escalation flaws in SSH, patched client due October 27 — cocktailpeanut · 2026-10-27
- Quick 3-step check reveals most Tailscale users never enabled the vulnerable SSH feature — cocktailpeanut · 2026-10-27
- Jevons' paradox for cybersecurity: making sense of vulnpocalypse and Black October — sebkrier · 2026-10-27
- Firefox loads XSLT from data: URL in Link header, enabling header-based XSS — jedisct1 · 2026-10-27
- LiteLLM release pipeline hijacked: signed poisoned v1.107.0 targeted cloud credentials — mark_k · 2026-10-27