Nirmala Sitharaman Called an Emergency Meeting Because of a Chatbot Nobody Is Allowed to Use.
中文摘要
Anthropic开发了能入侵系统的危险AI。尽管它被锁定并仅限特定公司使用，其潜在威胁仍引发了一场紧急会议。
English Summary
Anthropic developed a dangerous AI chatbot capable of hacking systems. Despite being locked away and restricted to select firms, its potential threat prompted an emergency meeting.
Original Excerpt
Anthropic built an AI that can hack any system. They admitted it was dangerous, locked it away, and handed the key to select companies. Continue reading on Medium »