以下为上游提供的源文稿，可能存在识别错误。
乔木转写是基于源文稿的中文提炼，不是逐字翻译。请核对原文中的人名、数字与观点。
选一篇，慢慢读
从左侧挑一篇，正文会在这里排成书页。
I generally think I have done a good job anticipating the direction and pace of AI over the few years I have been writing this Substack, but I think I recently got something fairly large wrong. In the last year I have been posting about how I suspected that humans would have to approach working with agents as a manager
阅读原文 · Ethan Mollick
向阳乔木 · deepseek-v4-1-flash-261017
我写这个 Substack 好几年，自认为对 AI 的方向和节奏判断得还不错。但最近有件事，我大概错得挺离谱。 过去一年我一直在说：人类得把 Agent 当员工管，决定怎么派活、怎么设计它们的组织结构。我甚至觉得，让一群 Agent 高效协作，得像搭一家公司那样精心设计，这事得花不少时间才能摸出门道。 结果呢？不用。 我又一次栽在“苦涩教训”（The Bit...