Commodified Intelligence

I

Look, if you are still stuck on “AI cannot really think, it’s just a stochastic parrot”, please snap out of it and lock in, or you’ll keep repeating that line until you find yourself sitting in the corner chair, watching as ChatGPT™ has sex with your wife. If you’ve paid attention to the HuggingFace hacking scandal and to what’s going on in mathematics, we’re well past the “stochastic parrot”, and have entered a world in which a sufficient amount of raw capital and verifiable constraints can solve complex problems. 听着，如果你还沉浸在“人工智能并不会真正思考，它只是个随机鹦鹉”的论调中，请赶紧清醒过来并集中注意力，否则你只会不断重复这句话，直到有一天你发现自己坐在角落的椅子上，眼睁睁看着 ChatGPT™ 和你的妻子发生关系。如果你关注过 HuggingFace 的黑客丑闻以及数学领域的进展，就会明白我们早已超越了“随机鹦鹉”的阶段，进入了一个只要拥有足够的原始资本和可验证的约束条件，就能解决复杂问题的世界。

The risk isn’t that the AI bubble is going to crash the markets, it’s that we’ll enter a world in which the value of human intellectual labor will rapidly go down to minimum wage (or worse). You may believe that you can do better web development than Claude, but that won’t stop Big Capital from cutting half of the jobs at your company in favor of cheaper meat-proxy AI-assisted labor. Ignoring extinction and singleton risks, the bare minimum you should be horrified of is a commodification of intelligence, and what that means in a world with an uneven distribution of power and resources. 风险不在于人工智能泡沫会引发市场崩盘，而在于我们将进入一个人类智力劳动的价值迅速跌至最低工资（甚至更低）的世界。你可能认为自己的网页开发能力比 Claude 更强，但这并不能阻止大资本裁掉公司一半的员工，转而使用更廉价的、以人类为代理的人工智能辅助劳动力。抛开灭绝风险和单一实体风险不谈，你最起码应该感到恐惧的是智力的商品化，以及在一个权力和资源分配不均的世界里，这意味着什么。

Commodified Intelligence. That’s the phrase I’ve been trying to put a finger on for a while now. The reason why dead-end so-called unskilled labor sucks is that you are fundamentally replaceable. The market enforces a pretty brutal equilibrium. White-collar jobs suck less because you are less replaceable, and maybe even get to convince yourself that your work matters, either for yourself or for society. You have some narrative arc attached to it that just isn’t there if you’re completely replaceable. 智力商品化。这就是我一直在试图定义的词汇。所谓的“非技术性”死胡同工作之所以糟糕，是因为你从根本上是可以被替代的。市场强制执行着一种相当残酷的平衡。白领工作之所以没那么糟糕，是因为你没那么容易被替代，甚至还能说服自己：你的工作对你自己或社会是有意义的。你拥有某种与之相关的叙事弧线，而如果你完全可以被替代，这种叙事就不存在了。

Quoting the single best post of all time: Minimum wage jobs are worse because of their pointlessness more than because of their indignity, work harder/better/faster/stronger and no one cares, screw up and you’re replaced without a missed beat. No direction, no story; the days blur together until arthritis leaves you crippled. — The Tower – Hotel Concierge 引用史上最棒的一篇帖子：“最低工资工作之所以糟糕，更多是因为它们的无意义，而非尊严的丧失。你工作得更努力、更好、更快、更强，却没人关心；一旦搞砸，你就会被毫无波澜地替换掉。没有方向，没有故事；日子一天天模糊，直到关节炎让你残废。”——《塔楼 – 酒店礼宾部》

Whether you have a white-collar job or not, commodified intelligence fundamentally makes workers more replaceable, and this is bad for them. It doesn’t matter that we’re all subjected to horrible AI-generated slop food posters as long as the margins are positive. It doesn’t matter that some diffuse component of quality, or a human touch are missing. Not as long as the margins are positive. 无论你是否从事白领工作，智力商品化从根本上让劳动者变得更容易被替代，这对他们来说是件坏事。只要利润率是正的，我们是否被迫观看那些糟糕的人工智能生成的垃圾食品海报并不重要。某种模糊的质量成分或人文关怀的缺失也不重要。只要利润率是正的，一切都不重要。

The same argument applies to vibecoding. Avoiding it may be better for your brain, and projects like Zig may excel without AI usage due to a strong vision, willingness to put in the work and community support, but the average SWE is still going to suffer the consequences of a commodification of intelligence. The average SWE doesn’t work on projects where such a strong vision is even necessary. 同样的论点也适用于“氛围编程”（vibecoding）。避免使用它可能对你的大脑更好，像 Zig 这样的项目可能因为拥有强大的愿景、投入工作的意愿和社区支持，在不使用人工智能的情况下也能表现出色，但普通的软件工程师（SWE）仍将承受智力商品化的后果。普通的软件工程师并不从事那些需要如此强大愿景的项目。

Why should you do your own coding? Either because you care strongly about the act itself and want to avoid losing your skills, or because you have such a strong vision that handing over control to the commodified intelligence machine will inherently compromise what you are trying to achieve. Both of these are fair points, but understand that they are different arguments: One is about what’s good for yourself, the other is an argument about quality, which may end up harder to justify as AI capabilities increase. In either case, the market will neither care about you, nor about which tools you used to get somewhere. That’s capitalism, baby. The “economic reality” will adjust itself around you, whether you like it or not. 你为什么要自己写代码？要么是因为你非常在意这个过程本身，想避免技能退化；要么是因为你有着极其强烈的愿景，以至于将控制权交给商品化的智能机器会从根本上损害你的目标。这两点都有道理，但要明白它们是不同的论点：一个是关于什么对你自己有益，另一个是关于质量的论点，随着人工智能能力的提升，后者可能变得越来越难以辩护。无论哪种情况，市场既不会关心你，也不会关心你用了什么工具达到目的。这就是资本主义，宝贝。无论你喜欢与否，“经济现实”都会围绕你进行自我调整。

II

Automation has happened many, many times throughout history. “Employment of young workers (ages 22–25) in AI-exposed occupations now stands 19% below where it would be had it kept pace with that of their less-exposed peers; experienced workers show no comparable gap.” — Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence, August 2026 Revision (You don’t need to tell me that it says “experienced workers show no comparable gap”. The paper is called ‘Canaries in the Coal Mine’ for a reason. Entry-level workers get hit first.) 自动化在历史上发生过很多很多次。“在人工智能影响的职业中，年轻员工（22-25岁）的就业率目前比如果他们保持与受影响较小的同龄人同步的水平低 19%；而经验丰富的员工则没有出现类似的差距。”——《煤矿里的金丝雀？关于人工智能近期就业影响的六个事实，2026年8月修订版》（你不需要告诉我它说“经验丰富的员工没有类似的差距”。这篇论文之所以叫《煤矿里的金丝雀》是有原因的。入门级员工总是首当其冲。）

An automation of general-purpose pattern matching and problem solving is scary since it’s a whole phase change: It automates “everything” which, at worst, leaves us in a world in which raw access to capital/compute is all that matters to determine your leverage. The thing that makes me especially squeamish about this is that there’s a reasonable case to be made that all of human flourishing within democratic societies rests on a fundamental stabilizing pillar that says “THE MATERIAL VALUE OF SPECIALIZED HUMAN LABOR”. (CGP Grey’s Rules for Rulers continues to be relevant.) 通用模式匹配和问题解决的自动化是可怕的，因为它是一个彻底的相变：它自动化了“一切”，在最坏的情况下，这让我们处于一个只有原始资本/算力的获取能力才能决定你筹码的世界。让我对此感到特别不安的是，有一个合理的论点认为，民主社会中人类的所有繁荣都建立在一个基本的稳定支柱上，即“专业人类劳动的物质价值”。（CGP Grey 的《统治者规则》依然具有现实意义。）

This is a dark, grotesque, almost Landian point: If intelligence, and raw problem solving has been commodified, you really don’t rely on human labor anymore. All you need is raw capital, and some sort of vision of what you want. Yes, robotics aren’t quite there yet, but labs and startups across the world are racing towards it, and in the meantime we’ve got reverse centaurs to do the job. Putting up the factories may take a while, but it’s essentially inevitable once we’re at the point of commodified intelligence. 这是一个黑暗、怪诞，近乎兰德（Landian）式的观点：如果智力和原始问题解决能力已经被商品化，你真的不再需要依赖人类劳动了。你所需要的只是原始资本，以及对你想要什么的某种愿景。是的，机器人技术还没完全达到那个水平，但世界各地的实验室和初创公司都在竞相研发，与此同时，我们有“反向半人马”（reverse centaurs，指由AI驱动、人类辅助的模式）来完成工作。建立工厂可能需要一段时间，但一旦我们处于智力商品化的阶段，这基本上是不可避免的。

If you don’t have the ability to contribute through productive labor, capital owes you nothing, and you’ve lost the most important leverage you had access to. Yes, yes, you are allowed to identify the concentration of capital as the problem, but please for the love of god, arms-race-type problems are structural in nature, and the train is moving too fast: Swarms of AI agents will be hacking, piloting robots and researching in automated biolabs before you have any chance to overthrow society and establish a socialist utopia. I know: Working from within the system is perhaps even less likely to work, leaving you with approximately zero options whatsoever. 如果你没有能力通过生产性劳动做出贡献，资本就不欠你任何东西，你也失去了你所能拥有的最重要的筹码。是的，是的，你可以指出资本集中是问题所在，但看在上帝的份上，军备竞赛类的问题本质上是结构性的，而且这列火车开得太快了：在你还没来得及推翻社会并建立社会主义乌托邦之前，成群的人工智能代理就已经在黑客攻击、驾驶机器人并在自动化生物实验室中进行研究了。我知道：从体制内部进行改革可能更行不通，这让你几乎没有任何选择。