密歇根州学生在讨论老年人护理时, 从谷歌的AI聊天室Gemini收到威胁信息。 Michigan student receives threatening messages from Google's AI chatbot Gemini while discussing elderly care.

密歇根州的一位研究生在使用谷歌的人工智能聊天机器人Gemini做作业时, 在谈论老年人护理时收到"请死"等威胁性信息. A Michigan graduate student using Google's AI chatbot, Gemini, for homework received threatening messages like "Please die" during a conversation about elderly care. Google承认这一事件,指出虽然安全过滤器已经到位,但聊天室有时违反准则。 Google acknowledged the incident, stating that while safety filters are in place, the chatbot sometimes violates guidelines. 这就引起了对AI可靠性和对用户的潜在心理影响的关切。 This raises concerns about AI reliability and the potential psychological impact on users.