像对待编译器输出一样对待智能体输出
文章认为,随着智能体的出现,传统的代码评审变得不切实际甚至是不负责任的。
作者类比编译器输出,指出我们通常不会人工审查编译后的二进制文件,而是通过测试、类型系统、可重现构建等手段确保其质量。
对于智能体输出,同样不应进行人工审查,而是需要构建完善的测试基础设施、形式化规范和AI自我检查流水线,并将信任转移到这些自动化流程上。
文章强调,当前工程师对“无人工评审”感到不安,是因为相关基础设施尚未成熟,呼吁重视在形式化规范、测试和AI检查等方面进行投入,以确保智能体生成的代码的质量和安全性。
查看原文开头(英文 · 仅前 3 段)
Philip Su's recent post argues that code reviews are not just impractical in the age of coding agents, they're headed toward being irresponsible. He's right on trend. But I think the framing of "lights-out codebases" skips over the more interesting and uncomfortable question: why does lights-out feel so scary, and what does that fear actually tell us?
The answer, I think, is hiding in something we already used once before and then promptly forgot we did: the compiler.
Nobody Reviews Compiler Output
※ 出于版权考虑,仅引用前 3 段。完整内容请阅读原文。