Safety advocate pushes back on Robb: if it isn't safe, don't deploy it — period
gerardsans · x · 2026-10-15
Responding to John Robb's three society-scale alignment approaches, Gerard Sans argues the framing itself is wrong: human safety must override capability or reach, always. He criticizes Anthropic for consistently treating safety as an afterthought priority — and states plainly that if a system isn't safe, it shouldn't be deployed at all.
Related event: Debate over AI alignment: human-machine pairing vs safety-first deployment(2 posts)→
More from Safety
- Florida AG asks court to bar OpenAI from training new models without third-party approval — imjustnewatai · 2026-10-23
- OpenAI to join Prompt2PWN 2026 AI hacking conference for live vuln research — OpenAIDevs · 2026-10-23
- Science lifts expression of concern on 2024 AI debunking-conspiracies paper; conclusions unchanged — DG_Rand · 2026-10-23
- Chamath: Banks will block AI agents via terms of service before anything else — nicolechirps · 2026-10-23
- Reddit thread argues safety research is indistinguishable from capabilities research — ResultBackground2450 · 2026-10-23
- Beam vs GLM 5.3 Abliterated: 10-round attack/defense test shows 96% of vulns patched — bigblueboo · 2026-10-23