Agents Don’t Escape Models. They Escape Architectures.
中文摘要
本文探讨了AI智能体，认为其突破不在于模型，而在于系统架构。
English Summary
The article argues that AI agents' capabilities are determined by their architectures rather than the underlying models.
Original Excerpt
Last Tuesday a CISO asked me the question I now refuse to answer. We were in a windowless room in Madrid, looking at an agent stack that… Continue reading on Medium »