AI Made Me 5x Faster. It Also Made Me 5x Worse at My Job.
AI Made Me 5x Faster. It Also Made Me 5x Worse at My Job.
AI 让我的工作效率提升了 5 倍，但也让我的工作质量下降了 5 倍。
There is a moment happening in offices and bedrooms all over the world right now, and it looks like this. Three AI agent sessions are running. One is refactoring something. One is writing tests. One is halfway through a migration nobody wanted to do by hand. The person in front of them is not typing. They are watching. Their eyes move between panes like someone who put chips on three tables and cannot decide which one to be nervous about. 此时此刻，在世界各地的办公室和卧室里，正发生着这样一幕：三个 AI 智能体（Agent）会话正在运行。一个在重构代码，一个在编写测试，另一个正在进行一项没人愿意手动完成的迁移工作。坐在屏幕前的人并没有在打字，而是在观察。他们的目光在不同的窗口间游移，就像一个在三张赌桌上都下了注，却不知道该为哪一张感到紧张的赌徒。
Then the thought arrives. I could start a fourth one. I call it the fourth terminal, and I think it is the defining mistake of the AI coding era. Not because running agents in parallel is bad. Because of what the reflex reveals. AI handed us spare capacity, and our first instinct was to fill it with more AI, instead of asking what the spare capacity was actually for. 接着，一个念头浮现：我可以再开第四个。我称之为“第四终端”，我认为这是 AI 编程时代最典型的错误。这并不是因为并行运行智能体本身不好，而是因为这种下意识的反应暴露了问题：AI 为我们提供了多余的产能，而我们的第一反应竟是用更多的 AI 去填满它，而不是去思考这些多余的产能究竟应该用来做什么。
I made that mistake for about four months. Nearly everyone I know made it too. This is what it cost, what the research now says about why it happens, and what the people who came out the other side are doing instead. 我犯了这个错误大约四个月。我认识的几乎每个人也都犯过。以下是这一行为的代价、研究对这一现象的解释，以及那些走出误区的人们现在正在做的事情。
The AI honeymoon is real, and you should enjoy it
AI 的蜜月期是真实的，你应该享受它
Let me be fair to the tools first, because the backlash has gotten lazy. When agentic coding properly landed in my workflow, it felt like someone lifted a weight off my chest I had not known I was carrying. All that configuration written by hand. All that run, squint, fix the typo, run again. Suddenly optional. 首先，我要对这些工具保持公正，因为现在的反弹情绪已经变得有些懒惰了。当智能体编程真正进入我的工作流时，感觉就像有人卸下了我从未察觉到的重担。那些曾经需要手写的配置，那些“运行、眯眼检查、修复拼写错误、再次运行”的繁琐过程，突然变得不再必要了。
Work that used to take a day was done before lunch. A migration I had avoided for a quarter got drafted in an afternoon. My manager noticed. My team noticed. 过去需要一天完成的工作，现在午饭前就能搞定。我拖了一个季度的迁移任务，一个下午就完成了草稿。我的经理注意到了，我的团队也注意到了。
This is not vibes. In a controlled study of developers building a simple HTTP server, the ones with an AI assistant finished noticeably faster. In a field experiment across thousands of developers, merged pull requests rose by roughly a quarter. If your work involves a lot of greenfield code or a lot of boilerplate, the AI speedup is real and it is not small. 这并非错觉。在一项关于开发者构建简单 HTTP 服务器的对照研究中，使用 AI 助手的开发者完成速度明显更快。在针对数千名开发者的实地实验中，合并的 Pull Request（PR）数量增加了约四分之一。如果你的工作涉及大量新项目代码或样板代码，AI 带来的提速是真实且显著的。
So we did the obvious thing. We got faster, so we took on more. Bug report that smells like infrastructure? I am on it. Someone needs a dashboard by Friday? Sure. Ticket from March rotting in the backlog? Why not. My open pull request count started to look like a typo. 于是我们做了显而易见的事：我们变快了，所以我们承担了更多。看起来像基础设施问题的 Bug 报告？我来处理。有人周五前需要一个仪表盘？没问题。三月份积压在任务列表里发霉的工单？为什么不呢。我待处理的 PR 数量开始多得像个输入错误。
The bill AI quietly runs up
AI 悄悄累积的账单
Here is what nobody tells you about being five times faster. You can also get lost five times faster. The first sign was easy to ignore. A colleague asked about one of my open PRs, and I had to read my own description to remember what it was for. The description had been written by AI. I was reading a machine’s summary of a decision I had apparently made, in order to find out what I thought. 关于效率提升 5 倍，没人会告诉你的是：你迷失方向的速度也会快 5 倍。第一个迹象很容易被忽略：一位同事问起我其中一个待处理的 PR，我不得不读一遍自己的描述才能想起它是做什么的。而那段描述是 AI 写的。我正在阅读机器对我所做决策的总结，以此来弄清楚我当时是怎么想的。
That is not productivity. That is a queue with your name on it. 那不是生产力，那只是一个挂着你名字的待办队列。
The second sign was not small. An agent produced a change adding a new permission set for a service. Clean diff. Sensible naming. Tests green. I reviewed it the way I had started reviewing everything by then, which is to say I scrolled, nodded, approved. It had been open six days and I wanted it gone. 第二个迹象就不那么小了。一个智能体生成了一个为服务添加新权限集的变更。差异对比很干净，命名合理，测试通过。我以当时审查一切的方式审查了它——即滚动页面、点头、批准。它已经挂了六天，我只想赶紧处理掉。
Two things saved me. A teammate who actually reads policy documents left one comment: “is this wildcard on purpose?” And luck, in that the comment landed before the merge did. The AI had done exactly what I asked. I had asked for the wrong thing, vaguely, and it filled the gap with the most permissive option available. The tests passed because the tests checked that the permission existed, not that it was safe. 两件事救了我。一位真正阅读策略文档的队友留下了一条评论：“这个通配符是故意的吗？”还有运气，因为这条评论在合并之前发出了。AI 完全按照我的要求做了。但我要求得不对，且很模糊，于是它用最宽松的选项填补了空白。测试通过了，因为测试只检查了权限是否存在，而不是检查它是否安全。
That is the sentence that now governs my working day: “All tests pass” is not the same as “this does what I meant.” Everything else in this article comes out of that gap. 这句话现在成了我工作的准则：“所有测试通过”并不等于“这实现了我的意图”。本文其余部分皆源于这一认知鸿沟。
AI did not remove the bottleneck. It moved it.
AI 没有消除瓶颈，它只是转移了瓶颈。
The most useful way I have found to think about this: the bottleneck did not disappear, it relocated. The expensive part of engineering used to be turning an idea into syntax. Everything we built was shaped around that assumption. How we estimate, how we review, how we staff, what a good day feels like. AI made generating a first draft nearly free. So the expensive part became two other things: saying precisely what you want, and proving you got it. 我发现思考这个问题最有用的方式是：瓶颈并没有消失，它只是换了个位置。工程中最昂贵的部分曾经是将想法转化为语法。我们构建的一切都是围绕这个假设形成的：我们如何估算、如何审查、如何配置人员、什么样才算充实的一天。AI 让生成初稿变得几乎零成本。因此，昂贵的部分变成了另外两件事：精确地表达你的需求，以及证明你得到了你想要的。
The evidence for that relocation is uncomfortable. On teams with high AI adoption, pull request review time has been measured growing by around 90%, in data drawn from over ten thousand developers. Review time nearly doubling is not a sign people became more careful. It is a sign the volume arriving at review outgrew the humans standing there. 这种转移的证据令人不安。在 AI 采用率高的团队中，根据超过一万名开发者的统计数据，PR 审查时间增加了约 90%。审查时间几乎翻倍并不是因为人们变得更谨慎了，而是因为送达审查的工作量超出了人类的处理能力。
There is a stranger finding too. One study of experienced open source developers working on their own large codebases found they were about 19% slower with AI, while believing they were 20% faster. That study was later revised over a sample selection problem, so treat it as one signal rather than a verdict. But the shape matches what a lot of senior engineers quietly report. On a big messy system you already know well, AI often is not saving your time. It is moving your time from writing to checking, and checking feels like less work even when it is not. 还有一个更奇怪的发现。一项针对在大型代码库上工作的资深开源开发者的研究发现，使用 AI 后，他们的速度反而慢了约 19%，但他们却认为自己快了 20%。该研究后来因样本选择问题进行了修订，所以请将其视为一个信号而非定论。但这种情况与许多资深工程师私下反映的情况相符。在一个你已经很熟悉的庞大且混乱的系统中，AI 往往并不能节省你的时间。它只是将你的时间从“编写”转移到了“检查”上，而检查工作即使实际上并不轻松，也会给人一种“工作量较小”的错觉。
Notice the condition, because this is where most online arguments go wrong. AI is fast on small and new. It is unreliable on large and old. Those are different situations producing different results, not two camps shouting past each other. 请注意这个前提，因为大多数网络争论都错在这里。AI 在处理小型和新项目时很快，但在处理大型和旧项目时不可靠。这是两种产生不同结果的情境，而不是两个阵营在互相叫嚣。
Then there is the AI trust paradox, which is almost funny. A large share of production code is now co-authored by AI, while roughly three-quarters of developers report frequently seeing hallucinations and say they do not fully trust the output. We are delegating more to something we trust less. 此外还有一个几乎有些滑稽的“AI 信任悖论”。目前很大一部分生产代码是由 AI 共同编写的，而大约四分之三的开发者表示经常看到 AI 产生幻觉，并称他们并不完全信任其输出。我们正在将更多的工作委托给一个我们信任度更低的东西。
The space between how much we delegate and how much we trust deserves a name: verification debt. Time saved on writing that comes back later as an audit of code nobody understands. Unlike ordinary technical debt, it never appears on a board. It appears at 2 a.m. 我们委托的程度与信任的程度之间的差距，值得起一个名字：验证债务（Verification debt）。在编写代码时节省的时间，后来会以“审计没人能看懂的代码”的形式加倍奉还。与普通的技术债务不同，它永远不会出现在任务看板上，它只会出现在凌晨两点。
Why the fourth terminal is so hard to resist
为什么“第四终端”如此难以抗拒
If this were only about tooling, we would have fixed it already. It is about what AI does to your sense of yourself. Three forces push you toward… 如果这仅仅是工具问题，我们早就解决了。这关乎 AI 对你的自我认知产生了什么影响。有三种力量在推动你走向……