告诉 HN： “少写代码” 现在有了两种不同的含义
1 分•作者: mikaelaast•7 天前
今天早上，我在Slack私信里收到了几张Twitter截图，让我感到困惑。其中一个是Boris Cherny说的，Claude Code的代码几乎100%都是由Claude Code自己编写的，而且他已经好几个月没有手动进行过哪怕是很小的编辑了。另一张截图是OpenAI的一位技术人员说的，编程“一直都很糟糕”，当被问及OpenAI模型完成了多少代码时，他回答说：“100%，我再也不写代码了。”
我对这种“摆脱它”的态度感到沮丧。这里其他人也有同样的感觉吗？
我的目标也是尽可能少地写代码。这并不是因为我想把工作推给AI，而是因为我希望代码越少越好。代码就意味着责任，而bug就潜伏在代码里，所以我总是力求做到最少。
我会在脑海中建立一个代码库的模型。在我的头脑中，它被抽象成一个“思维宫殿”，在那里我可以自由移动，不受语言或语法的限制。最近，我开始将代码分成两类：一类是我不需要在脑海中建模的代码（低风险，因为它遵循既定的规范），另一类是我无法抗拒在脑海中建模的代码（创新，新模式，高风险）。第一类很容易委托给AI代理。第二类则太有趣、太令人满意，让人无法抗拒。我会和AI代理一起切磋，但我仍然希望我的思维宫殿与代码保持同步。
我们真的要放弃我们与代码的关系吗？
查看原文
I was the bemused recipient of a couple of Twitter screenshots in a Slack DM this morning. One was Boris Cherny saying pretty much 100% of Claude Code's code is written by Claude Code, and that he personally hadn’t made even small edits by hand for months. The other was a Member of Technical Staff at OpenAI saying programming "always sucked" and when asked what percentage of coding is done from OpenAI models, replied: "100%, I don’t write code anymore."<p>I'm disheartened by this "good riddance" attitude. Do others here feel the same?<p>My goal is also to write as little code as possible. Not because I want to push the work onto AI, but because I want as little code to exist as possible. Code equals liability, and bugs live in code, so I always strive for the bare minimum of it.<p>I keep a mental model of a codebase. In my head it’s abstracted into a kind of "mind palace" where I can move around without language or syntax. Lately I’ve started separating code into two categories: code I don’t need to model in my head (low liability, because it follows established conventions), and code I can’t resist modeling in my head (innovative, new patterns, high liability). The first is easy to delegate to an AI agent. The second is too enjoyable and satisfying to stay away from. I’ll spar with an agent, but I still want to keep my mind palace in sync with the code.<p>Are we really going to give up our relationship with code?