AI Safety Should Pivot to Resilience Over Guardrails
joshua_saxe · x · 2026-07-18
The core of this discussion argues that AI safety has become incoherent as a research program, but this doesn't mean we should default to laissez-faire accelerationism. The author advocates for realigning the safety agenda:
- Decrease the hyper-focus on jailbreaks and traditional guardrails, as models will proliferate uncoordinatedly, making unified guardrails unreliable in many scenarios.
- Increase the emphasis on 'resilience,' focusing on how systems and society can withstand risks in an environment where models are widely accessible.
- De-emphasize purely thought-experiment-based risk arguments, without ignoring them entirely.
Overall, this represents a paradigm shift from 'locking down models' to 'building resilient governance facing the reality of proliferation.'
More from AGI Musings
- All 8 Transformer paper authors left Google, and rival labs captured most of the frontier value — signulll · 2026-10-27
- Anthropic's S-1 pledges AI for curing diseases and serving 4.6B underserved patients — petergyang · 2026-10-27
- Goldman finds statistically significant negative link between AI adoption and US employment growth — ben_j_todd · 2026-10-27
- a16z: World-changing tech makes a chaotic entrance — AI agents are following the same path — a16z · 2026-10-27
- NEJM AI on medical education: mastery requires performing without the machine — BraydonDymm · 2026-10-27
- Dev warns abliterated models could wreck software supply chains with 1000+ transitive dependencies — nitarshan · 2026-10-27