AI safety organizations have long worked on cybersecurity and misuse, push back on critics
panickssery · x · 2026-09-22
A thread pushes back on the popular claim that AI safety/EAs should work on cybersecurity, model theft, and constraining misbehavior instead of alignment — arguing these have in fact been major focuses of leading organizations in the space for years.
More from AGI Musings
- AI safety researcher David Krueger calls for immediate international moratorium on frontier AI — DavidSKrueger · 2026-09-22
- Shopify CEO Tobi Lütke: the one thing AI can't do is take responsibility — andreisavu · 2026-09-22
- Hinton: Superintelligence Alignment Should Be Trained In, Not Locked Down — victor_explore · 2026-09-22
- Hundreds of Studies: Work's Non-Financial Benefits Can Come From Outside Traditional Jobs — scychan_brains · 2026-09-22
- Paper Review: UBI Alone Is Insufficient for Wellbeing Under AI Labor Displacement — scychan_brains · 2026-09-22
- Financial Dependence Undermines Agency: Gulf States Offer a Cautionary Tale for AI Futures — scychan_brains · 2026-09-22