← Back to Library

What would it look like if the AI bubble popped?

Matt Stoller cuts through the hype of artificial intelligence to ask a question that financial markets are desperate to avoid: what happens when the subsidy stops? While Wall Street fixates on the transformative potential of algorithms, Stoller argues we are witnessing a classic speculative mania where stock valuations have completely detached from economic reality. This piece is vital because it shifts the conversation from "will AI change the world?" to "what breaks first when the money runs out?"

The Disconnect Between Price and Value

Stoller begins by dismantling the assumption that current market highs reflect genuine productivity. He notes a consistent theme in modern finance: "the stock market is increasingly disconnected from underlying economic activity." This isn't just a minor inefficiency; it's a structural feature of an economy driven by the "Number Go Up Rule," where policy and corporate strategy are optimized solely to inflate capitalization.

What would it look like if the AI bubble popped?

The author draws a sharp line between the technology itself and the financial bubble surrounding it. "It's hard to offer persuasive tech policy arguments in the midst of a bubble," Stoller writes, suggesting that governance only happens in crisis. This is a crucial distinction for busy readers: we are currently debating the ethics of AI while ignoring the fact that its financing model is fundamentally fragile. The core argument rests on the idea that "the speculative discourse only works as long as investors subsidize the use of the technology." Once that subsidy erodes, firms must deliver measurable returns or face a reckoning.

When that subsidy stops, these AI firms have to actually deliver value, or customers won't buy it.

This framing is effective because it grounds the abstract concept of "AI" in hard economic constraints: electricity costs and hardware expenses. Stoller points out that just seven stocks linked to AI now comprise a third of the entire stock market, creating an economy-sized bet on enterprises that may not be profitable. Critics might argue that true transformative technologies often look unprofitable for years before scaling, but Stoller's comparison to previous manias suggests this time is different because the scale of investment—between $750 billion and a trillion dollars—is already distorting macro-economic growth numbers.

The Mechanics of Contagion

The piece moves from valuation to the mechanics of how a crash would spread. Stoller warns that "when a sector leading our financial markets craters, the problem escapes that sector and turns into a broad-based crash." He introduces the concept of contagion, explaining how balance sheets transmit shocks across the economy. The upcoming SpaceX IPO serves as a timely example; as universities and early investors gain paper wealth, they become dependent on those valuations. If the AI bubble bursts, these institutions will face immediate pressure to restructure their balance sheets, creating knock-on effects throughout the financial system.

A fall back to long-term average would cut, according to Baker, $300,000 per household of paper wealth from balance sheets.

Stoller acknowledges a counterargument: some economists claim high valuations are justified by a shift in income distribution rather than speculation. They argue that if corporations invest less and pay workers less, the resulting free cash flow for buybacks and dividends keeps stock prices high. However, Stoller counters this by noting that "now data centers are eating up all that cash flow." The argument holds weight because it highlights how the current boom is consuming the very capital that previously supported stable valuations.

2008 vs. The Dot-Com Era

To understand the potential shape of a crash, Stoller compares the AI bubble to two historical precedents: the 2008 financial crisis and the dot-com bust of 2000. He argues that 2008 is the wrong analogy because that crisis was driven by massive private debt and leverage—specifically mortgage-backed securities. "Speculating with other people's money creates a boom on the way up, but introduces significant fragility into the system," he explains. Today, while there is concern about "private credit" in the AI sector, the overall debt load is more balanced between government and private entities compared to 2007.

Unlike 2008, this bubble was not a surprise. Greenspan himself noted 'irrational exuberance' in 1996, four years before it popped.

Instead, Stoller posits that the dot-com era is the better mirror. The mechanism then was over-investment in infrastructure—fiber-optic networks—that turned out to be largely fraudulent or wasteful, much like the current data center build-out. "The boom in fiber, in other words, was a result of fraud," he notes regarding WorldCom, but adds that the broader sector saw massive speculative capital inflows. The parallel is striking: just as the dot-com crash led to a recession centered on tech and manufacturing before the Fed engineered a housing boom to replace it, an AI collapse could trigger an "investment-led recession."

The AI data center build-out is fostering significant economic growth... If they collapse, we could see an investment-led recession.

This comparison adds necessary depth by referencing the 1995 Private Securities Litigation Reform Act, which made it easier for CEOs to issue misleading statements—a legal environment that mirrors today's regulatory gaps. It suggests that the solution to a burst bubble might not be a bailout of homeowners, as in 2008, but rather a desperate attempt by policymakers to engineer a new speculative boom elsewhere to fill the void.

Bottom Line

Stoller's most compelling insight is that the AI bubble is not just a tech story, but a macroeconomic event driven by the same forces that created the dot-com crash: over-investment in infrastructure fueled by speculative capital. The piece's greatest strength is its refusal to get bogged down in the moral philosophy of AI, focusing instead on the hard math of financing and balance sheet contagion. However, it leaves open the question of whether the government will learn from history or simply repeat the cycle of replacing one bubble with another to avoid a recession. Watch for the moment when corporate earnings reports fail to justify the trillion-dollar infrastructure spend—that is the true trigger point.

Deep Dives

Explore these related deep dives:

  • The Age of Surveillance Capitalism Amazon · Better World Books by Shoshana Zuboff

    How tech companies turned human experience into raw material for prediction and control.

  • Dot-com bubble

    The article explicitly selects this era as the primary analogy for understanding how a technology-driven valuation disconnect can resolve without necessarily triggering a systemic banking collapse.

  • Goodhart's law

    This economic principle explains the specific mechanism of 'governance in crisis' described by the author, where financial metrics become targets that are gamed until they cease to be useful indicators of real value.

  • Free cash flow

    Understanding this specific phase of business investment is crucial to grasping why the article argues current AI spending represents a 'subsidy' that must eventually transition into measurable revenue or face a sharp correction.

Sources

What would it look like if the AI bubble popped?

One of the consistent themes of this newsletter is how the stock market is increasingly disconnected from underlying economic activity. In the Number Go Up Rule, I traced how we increasingly run everything to ensure that market capitalization continues to increase. From the dot com boom to the subprime housing to crypto to GameStop to sports gambling, there’s an increasing mania in how we encourage speculation instead of morally valuable activity.

The flip side of this disconnect is that governance happens in crisis. Our collective understanding of finance and politics is shaped by crashes - the dot com boom, the Great Financial Crisis of 2008 and Covid, all episodes in which a crisis in one part of the system led to seemingly uncorrelated shocks elsewhere, and then political action to reorder the economy.

Today I want to ask what the popping of the AI bubble would look like, and whether it would precipitate a broader financial crash. And if it does so, what shape will it take? The right way to start is by analogy, as there are lessons from previous crashes and the governance that came out of them that we can learn from. For reasons I’ll get into, while 2008 and Covid could be useful to look at, the best analogy is dot com era.

First, it’s important to scope out what I’m not going to talk about, which is the governance of AI as a technology. It certainly matters whether we are creating a God-like system, a useful general purpose technology, or a moderately useful toy. There are many fascinating questions around copyright, liability, monopolization, and so forth, but it’s hard to offer persuasive tech policy arguments in the midst of a bubble.

In many ways, the key important question facing us today is the financing of AI, and the fact that we have placed a economy-sized bet on the enterprises claiming to focus on this technology. Just seven stocks - all linked to AI - comprise a third of the stock market, and AI capital investment is likely to be between $750 billion and a trillion dollars this year, which is big enough that it affects macro-economic growth numbers.

The stock mania we’re seeing as a result is based on the narrative that AI will be some sort of insanely profitable transformative technology. But AI is actually costly to operate, taking up a lot of electricity ...