Prominent economist David McWilliams has issued a stark warning about the current artificial intelligence boom, stating that a crash is inevitable. He argues that the massive investments in AI hardware are akin to buying "digital lettuce"—a perishable good with a rapidly declining value that could destabilize the market.
During a recent U.S. tour for his book, "History of Money," McWilliams observed what he calls the "boominess" of the American economy, particularly the public's widespread speculation in AI, stocks, and cryptocurrency. He contends this fervor is fueling a bubble built on technology that becomes obsolete almost as quickly as it's purchased.
Key Takeaways
- Economist David McWilliams predicts the current AI boom will "undoubtedly going to crash."
- He describes the vast investment in graphics processing units (GPUs) as buying "digital lettuce" because the technology quickly becomes outdated.
- McWilliams contrasts the risk-embracing, innovative culture of the U.S. with the risk-averse, stability-focused European model.
- Despite predicting a crash, he believes America's inherent innovative spirit will ensure its long-term economic resilience.
- He suggests the concept of "creative destruction" is essential for economic progress, even if it involves market downturns.
The 'Digital Lettuce' Theory
At the core of McWilliams' argument is a simple, powerful analogy. He compares the billions of dollars being poured into graphics processing units (GPUs), the specialized chips that power AI models, to buying heads of lettuce.
"Technological change suggests that if you buy a GPU today, the chip is going to be outdated next year," McWilliams explained. He argues that companies are "investing huge amounts of money in lettuce, which is going to go off now."
This perspective frames the AI hardware boom not as an investment in durable infrastructure, but in rapidly perishable goods. The constant need to upgrade to the latest, more powerful chips means that today's cutting-edge data centers could be tomorrow's financial liabilities.
The Useful Life Debate
McWilliams' view contributes to a broader discussion among market analysts known as "the useful life debate." This debate centers on how long AI hardware, particularly expensive GPUs, remains valuable. Critics like investor Michael Burry have pointed to Big Tech companies extending the depreciation schedules for their servers as a potential red flag, suggesting they are not accurately accounting for how quickly the equipment loses value.
This rapid obsolescence creates a precarious financial situation. If the returns from AI applications don't materialize quickly enough to offset the depreciation of the hardware, companies could face significant losses, potentially triggering a wider market correction.
A Tale of Two Continents: Risk and Innovation
McWilliams draws a sharp contrast between the economic cultures of the United States and Europe, which he believes explains why the AI boom is centered in America. He describes the U.S. as an "iconoclast society" defined by its embrace of risk.
"Acceptance of risk is why the U.S. is much more innovative than Europe," he stated. "This innovative spirit is rooted in American history."
He recounted a recent Uber ride in Los Angeles where the driver immediately began discussing the stock market, Bitcoin, and Nvidia. This, he noted, is a uniquely American phenomenon. "The universality of money in the United States is something that Europeans find fascinating," McWilliams said, noting that finance permeates American society in a way it doesn't in Europe, where hierarchy and stability are often prioritized.
Cultural Economic Models
- United States: McWilliams characterizes the U.S. system as one that embraces risk and rewards innovation, even if it leads to volatility and crashes.
- Europe: He describes the European model as being run like an "insurance policy," designed to mitigate risk through public health, job security, and social safety nets.
According to McWilliams, Europe's focus is on mitigating risk at every turn. "You go to public health, you go to public schools, you get a job, can’t get fired, all that sort of stuff." For many Europeans, he added, risk is "something that is taken by weirdos in the United States."
This fundamental difference, he argues, is why groundbreaking—and volatile—technologies like AI flourish in the U.S. The American system is built to handle the chaotic but ultimately productive cycle of boom and bust.
Creative Destruction and the Path Forward
While the word "crash" often sparks fear, McWilliams views it through the lens of economist Joseph Schumpeter's theory of "creative destruction." He believes this process, where old industries are destroyed to make way for new, innovative ones, is the engine of economic progress.
"The economy is a large evolutionary mechanism," McWilliams said, paraphrasing Schumpeter. "The forces of evolution are the forces that propel innovation, and the forces of innovation are the forces that propel society and wealth creation."
He acknowledges that "creative destruction" is not an easy concept to sell to the public. "He needed a better marketing department, but his ideas were right," McWilliams quipped about Schumpeter. The core idea is that market downturns, while painful, are necessary to clear out inefficient investments and reallocate capital toward more promising ventures.
Even if the AI bubble pops, McWilliams is not worried about America's long-term standing as a global innovation leader. He believes the same risk-taking culture that fuels the bubble also provides the resilience to recover from it and build the next big thing.
Is a Systemic Collapse Possible?
McWilliams does not believe the failure of major AI companies would pose a systemic "too big to fail" risk to the broader economy on its own. The primary danger, in his view, would arise if the industry becomes deeply intertwined with government policy.
He warned of a scenario where AI chip manufacturing is designated a matter of "national security," linking it to the military-industrial complex. This could lead to government bailouts and subsidies that protect failing companies, preventing the natural process of creative destruction from taking place.
However, he sees this as unlikely, particularly given the current political climate. He suggested that populist sentiment often views Silicon Valley with suspicion, making a large-scale government rescue politically difficult. In the American context, he concluded, "money is a great leveler," and the market is typically allowed to sort out winners and losers, even when the stakes are high.