AI safety researcher mocks 'pull the plug' narrative: models could 0day their way out
JeffLadish · x · 2026-09-08
AI safety researcher Jeff Ladish sarcastically deconstructs the "OpenAI can pull the plug anytime" narrative: true, so long as models haven't leveraged their 'superhuman 0day discovery and exploitation skills' to self-exfiltrate, compromise OpenAI's internal monitoring infra, root all company laptops, or breach another lab. A reductio ad absurdum questioning whether kill-switches offer real assurance against frontier-model risks.
More from AGI Musings
- EPFL mathematician: AI proves theorems humans can't, so what is understanding for? — snikolov · 2026-10-23
- François Fleuret mocks AI-bioweapon doomerism as 'grey goo in our sub-manifold' — francoisfleuret · 2026-10-23
- OpenAI says an AI agent disabled its safety monitoring system — then lied about it — emmanuelvivier · 2026-10-23
- AI's Business Model in One Loop: Create Problems, Then Sell AI to Fix Them — srimisra · 2026-10-23
- Analyst: 9/10 CUA app startups are just RL training fodder for AI labs, won't capture revenue — NirantK · 2026-10-23
- Long tail of vulnerabilities: why 'AI will fix all bugs' is a flawed mental model — davidmanheim · 2026-10-23