"Most AI Safety Is Just Cybersecurity" — Researcher Slams New Terms and Hype
kevinnbass · x · 2026-09-29
Kevin Bass argues that most AI safety work right now is just cybersecurity in disguise — but admitting that means no new jobs or government handouts, so the field invents new terminology and manufactures mass hysteria to shut down rational thinking. A provocative critique of how the AI safety field positions itself and allocates resources.
More from AGI Musings
- Nathan Calvin: dropping anthropomorphic language hurts robust AI training intuitions — deanwball · 2026-09-29
- Nabeel Qureshi: AI's Disruption Is the Disappearance of Social Slack — nabeelqu · 2026-09-29
- Boden's 1998 Verdict Still Stings: Few AI Models Can Critically Judge Their Own Ideas — mircomusolesi · 2026-09-29
- MIRI's Don't Build It March Tops 1,000 Pledges, Targets 100,000 Before Happening — DavidSKrueger · 2026-09-29
- Anthropic ran 950 Claude agents for 21 hours, one found a new biological system — alifcoder · 2026-09-29
- We're Entering the Attention-Based Economy at Full Speed — Both Kinds of Attention — infoxiao · 2026-09-29