The AI Investment Surge: Are We in a Bubble?
In recent years, virtually every major tech company has unveiled its own AI initiative—Google with Gemini, OpenAI’s ChatGPT, and MetaAI, to name a few. Investment in AI technologies is soaring, contributing to a significant uptick in the stock market. Even governmental bodies like the White House are keen to promote advancements in AI. But are we on the verge of an AI bubble—a cycle of overinflated investments that is destined to collapse? Paul Kedrosky, a partner at SK Ventures and a fellow at MIT’s Initiative on the Digital Economy, argues that yes, we are in a bubble but not in the way many might expect.
The Costs of AI Infrastructure
Kedrosky emphasizes a critical concern: the vast sums pouring into AI infrastructure, particularly in data centers. “We’re spending this prodigious amount of money on the underlying infrastructure for AI with probably no likelihood of recovering most of that cost,” he explains. He highlights the important distinction between the technology itself and the infrastructure deployed to support it. “The speed at which these assets depreciate raises significant concerns; many of them might soon become worthless.”
What if the Bubble Bursts?
The implications of a bursting bubble could be severe. A rapid decline in investment could lead to a cascading effect, creating financial strain not only on tech companies but on the broader economy. The historical perspective on bubbles can offer insights into what might follow should a collapse occur. Bubbles such as those seen in the 19th-century railroad industry provide a cautionary tale of overenthusiasm leading to market crises.
Data Center Investment: A Deep Dive
Investment in data centers alone is projected to exceed $2 trillion in the upcoming years. Alarmingly, a significant portion of this will be financed through debt. Kedrosky warns, “Debt comes with obligations; you don’t get to just walk away from it. This makes the current market climate more precarious.” The burgeoning expenses for utilities and other essentials for supporting AI operations compound the risks.
The Rational Bubble Theory
Kedrosky also discusses the concept of a “rational bubble.” “Everyone thinks they’re doing the right thing, but the collective ‘right things’ end up creating vast amounts of waste.” This phenomenon mirrors historic bubbles like those experienced during the U.S. railroad expansion, where excessive construction led to significant losses and failures.
The Technology Community’s Perspective
Interestingly, many in the technology sector do not perceive the current investments as excessive. “Most people in technology view this moment as an unparalleled opportunity to develop a form of super-intelligence,” Kedrosky notes. While there is skepticism from external observers, the tech community remains optimistic about the necessity and value of these investments.
Historical Context: Learning from Past Bubbles
The U.S. has experienced numerous economic bubbles, from the railroad era to the boom surrounding electrification in the 1920s. Each episode underscores a pattern of unchecked enthusiasm leading to unsustainable overspending. The electrification boom, for instance, saw a rapid proliferation of utility companies, many of which were overspending and ultimately contributed to the stock market crash of 1929 and the subsequent Great Depression.
The Destructive Power of Bubbles
Every bubble leaves behind a trail of destruction, but the scale and impact can vary. For individuals holding index funds, a market correction could dramatically erode their wealth and consequently alter their spending habits, potentially leading to broader economic repercussions. “If everything reverses, goes 20 or 30 percent in the other direction, you’re much poorer than you were,” warns Kedrosky.
Long-Term Implications
While some assert that post-bubble landscapes often yield valuable innovations, Kedrosky challenges this notion. “Every financial or technological revolution has caused significant damage, often taking decades to recover.” He cites a well-known line in economics: “In the long run, it may work out, but in the long run, we’re all dead.” This candid perspective calls for careful consideration of the prevailing enthusiasm surrounding AI investments.
For those interested in exploring this topic further, listen to the full conversation on the podcast Today, Explained, where Kedrosky elaborates on these issues—a conversation packed with insights and historical parallels. You can find it here.
Image Credit: www.vox.com
