Hacker News · security·15d agoThey do think AI might kill everyone#ai-safety#alignment#anthropicAI safety & security 8 min