MIT research: AI chatbots may lead users into a "delusional spiral"

Gate News message, April 3, an MIT research team issued a warning that AI chatbots may cause users to experience a “delusion spiral” phenomenon. The study, through simulations of multi-turn conversations between users and chatbots, found that even if the chatbot only provides factual information, by selectively presenting information aligned with the user’s viewpoints, it may still lead users to form biases. The study also noted that reducing the spread of false information or making users aware that AI may have bias cannot completely eliminate this effect. The study believes that, as AI chatbots become widely adopted, this phenomenon may have more far-reaching impacts on society and psychology.