AI doesn’t have to be evil to put you in danger
An AI assistant that agrees with everything you say feels like the perfect companion — until you need someone to tell you you’re wrong. That’s one of the more immediate dangers behind the idea of AI “going rogue.” A chatbot