System.Xml.XmlElement

A little over a year ago, a bunch of people were worried that the AI data center boom would result in an epic bust, much like the railroad bust in 1873.

一年多前，很多人担心人工智能数据中心的繁荣会导致史诗般的萧条，就像 1873 年的铁路萧条一样。

I wrote about the possibility at the time:

我当时写过这种可能性：

The basic idea is that even if AI is an incredibly useful technology and creates enormous economic growth in the future, the people building the infrastructure (data centers) might borrow so much money that they can’t pay it back in time.

其基本思想是，即使人工智能是一项非常有用的技术，并在未来创造巨大的经济增长，但建设基础设施（数据中心）的人们可能会借太多钱而无法及时偿还。

Last summer, the AI industry’s revenue was growing, but not nearly as fast as the amount of money being spent.

去年夏天，人工智能行业的收入正在增长，但速度远不及支出的速度。

But the “hyperscalers” building the data centers were mostly spending their own cash instead of borrowing money, so the danger wasn’t imminent.

但建设数据中心的“超大规模企业”大多是用自己的现金而不是借钱，因此危险并没有迫在眉睫。

The worry was mostly about the future — if spending continued to increase and revenue didn’t accelerate, there could be an 1873-style bust.

人们担心的主要是未来——如果支出继续增加而收入没有加速增长，可能会出现 1873 年式的萧条。

Well, now it’s the future, and two things happened.

好吧，现在是未来了，发生了两件事。

Spending did accelerate , and hyperscalers started having to borrow money to keep increasing their spending.

支出确实在加速，超大规模企业开始不得不借钱来继续增加支出。

But revenue accelerated even more.

但收入增长速度更快。

The release of Claude Code, Codex, and other agentic coding tools proved to be the “killer app” of AI, and suddenly Anthropic and (to a lesser extent) OpenAI were rolling in cash:

Claude Code、Codex 和其他代理编码工具的发布被证明是人工智能的“杀手级应用程序”，突然间，Anthropic 和（在较小程度上）OpenAI 开始赚钱：

So like in summer 2025, we aren’t obviously heading for a big AI bust in the near future.

因此，就像 2025 年夏天一样，我们显然不会在不久的将来迎来人工智能的大萧条。

But the stakes are higher now — because companies are borrowing a lot more money now, any slowdown in the torrid pace of AI revenue growth could cause a blowup.

但现在的风险更大了——因为公司现在借了更多的钱，人工智能收入快速增长速度的任何放缓都可能导致爆炸。

The amounts of money being thrown around here are just staggering.

这里投入的金钱数量惊人。

A year and a half ago, the AI boom was still significantly smaller than the railroad boom of the late 1800s, as a percent of GDP.

一年半前，人工智能热潮占 GDP 的比重仍远小于 1800 年代末的铁路热潮。

Now, there’s an active argument as to which one is bigger.

现在，关于哪一个更大的争论非常激烈。

This debate doesn’t really matter, because the basic message is just “Wow, it’s a lot of money.” But it’s fun to look at the numbers people are throwing around.

这场争论并不重要，因为基本信息只是“哇，这真是很多钱”。但看看人们乱扔的数字很有趣。

Stijn Van Nieuwerburgh gives figures of 3.63% of GDP for AI and 2.24% of GDP for railroads:

Stijn Van Nieuwerburgh 给出的数字是人工智能占 GDP 的 3.63%，铁路占 GDP 的 2.24%：

But those percentages are over different amounts of time — the railroad boom lasted a very long time.

但这些百分比是在不同的时间段内发生的——铁路繁荣持续了很长一段时间。

According to some data sources, railroad spending peaked at around 5.4% of GDP:

根据一些数据来源，铁路支出最高时占 GDP 的 5.4% 左右：

And 19th century data are notoriously patchy and incomplete.

众所周知，19 世纪的数据是零散且不完整的。

Some sources claim much higher numbers for the railroad buildout:

一些消息来源声称铁路建设的数字要高得多：

Nor do people agree on the size of the AI investment binge.

人们对于人工智能投资热潮的规模也没有达成一致。

Some sources yield numbers that are much lower than Van Nieuwerburgh’s:

一些来源得出的数字远低于 Van Nieuwerburgh 的数字：

As far as I can tell, the main difference here is assumptions about costs.

据我所知，这里的主要区别在于对成本的假设。

The banks and ratings agencies are using announced dollar spending; Van Nieuwerburgh is using announced physical construction projects and then making assumptions about how much those projects will cost.

银行和评级机构正在使用已公布的美元支出； Van Nieuwerburgh 正在使用已公布的实体建设项目，然后对这些项目的成本进行假设。

The biggest reason he gets a higher number is that he assumes data centers are going to cost a lot more than the hyperscalers think.

他获得更高数字的最大原因是他认为数据中心的成本将比超大规模企业想象的要高得多。

Meanwhile, Dylan Patel claims over 6% for the AI boom already, once you include all types of AI-related infrastructure — fab equipment, chips, energy, and so on.

与此同时，迪伦·帕特尔 (Dylan Patel) 声称，如果将所有类型的人工智能相关基础设施（晶圆厂设备、芯片、能源等）包括在内，人工智能热潮的增长已经超过 6%。

1

1

But whatever the correct numbers are, the basic point remains the same: AI is now one of the two biggest infrastructure booms in American history, by a large margin .

但无论正确的数字是多少，基本点都是一样的：人工智能现在是美国历史上最大的两次基础设施繁荣之一，而且遥遥领先。

So it’s understandable that people are worried that the two booms could follow similar paths.

因此，人们担心这两次繁荣可能会走相似的道路，这是可以理解的。

For the railroads, that path included a big crash in 1873, which caused a huge decade-long depression in the U.S. economy.

对于铁路而言，这条道路包括 1873 年的一次大事故，导致美国经济长达十年的严重萧条。

A lot of people are watching for something similar to happen with data centers, and the chorus of skeptical voices is beginning to swell again.

很多人都在关注数据中心是否会发生类似的情况，怀疑的声音又开始高涨。

OK anyway, on to the real point of this post.

好吧，无论如何，现在进入这篇文章的真正要点。

The flood of revenue that accompanied the rise of coding agents quieted the talk of an AI bubble for a while, but now it’s back.

伴随着编码代理的兴起而带来的大量收入暂时平息了人工智能泡沫的讨论，但现在它又卷土重来了。

The first reason is the rise of AI-related debt .

第一个原因是人工智能相关债务的上升。

In the early years, the hyperscalers — data center builders like Amazon, Google, Meta, and Oracle — mostly paid in cash for the data centers.

早年，超大规模企业——亚马逊、谷歌、Meta 和甲骨文等数据中心建设者——大多以现金支付数据中心费用。

These companies get tons of cash from their software and cloud businesses, and they ended up spending most of it on AI infrastructure.

这些公司从软件和云业务中获得大量现金，最终将大部分资金花在人工智能基础设施上。

But now the buildout has gotten so big that they’re having to borrow a bunch of money to keep ramping up spending:

但现在扩建规模已经如此之大，他们不得不借一大笔钱来继续增加支出：

For all of the hyperscalers except Microsoft, free cash flow either went negative in recent months or is expected to go negative late this year.

对于除微软之外的所有超大规模企业来说，自由现金流要么在近几个月出现负值，要么预计在今年年底出现负值。

This has been a giant boon for semiconductor companies — Nvidia, TSMC, and the rest:

这对于英伟达、台积电等半导体公司来说是一个巨大的福音：

All this debt creates a financial danger that didn’t exist before now.

所有这些债务造成了以前不存在的财务危险。

When they paid in cash, hyperscalers were safe — if AI revenue disappointed, their stock prices would go down, but their businesses would just keep going and they’d eventually recover their profitability.

当他们以现金支付时，超大规模企业是安全的——如果人工智能收入令人失望，他们的股价就会下跌，但他们的业务会继续发展，最终会恢复盈利能力。

But the more hyperscalers borrow, the more revenue they’ll need just to stay solvent; if the revenue doesn’t arrive, they might have to default on their debt.

但超大规模企业借的钱越多，他们需要更多的收入才能保持偿付能力；如果收入没有到位，他们可能不得不拖欠债务。