GPTProto

Dwarkesh Patels’s wildly popular but dangerously misleading account of the OpenAI Hugging Face incident

Gary Marcus:The Road to AI We Can Trust(RSS)·Aug 31, 2026, 11:24 PM·OpenAI / ChatGPT

Dwarkesh Patel 对 OpenAI/Hugging Face 事件的爆款解读被指危险地误导大众。Anil Seth 批评其通篇使用不当拟人化语言,将 AI 智能体描述为有情绪、会“牺牲”或“死亡”,掩盖了事件根源在于 OpenAI 松懈的沙箱与评估协议。