Photo by Alexas_Fotos on Unsplash
Some people think artificial intelligence will be the most important technology of the 21st century. Others insist that it is an obvious economic bubble. I believe both sides are right. Like the 19th century railroads and the 20th century broadband Internet build-out, AI will rise first, crash second, and eventually change the world.
The numbers just don’t make sense. Tech companies are projected to spend about $400 billion this year on infrastructure to train and operate AI models. By nominal dollar sums, that is more than any group of firms has ever spent to do just about anything. The Apollo program allocated about $300 billion in inflation-adjusted dollars to get America to the moon between the early 1960s and the early 1970s. The AI buildout requires companies to collectively fund a new Apollo program, not every 10 years, but every 10 months.
It’s not clear that firms are prepared to earn back the investment, and yet by their own testimony, they’re just going to keep spending, anyway. Total AI capital expenditures in the U.S. are projected to exceed $500 billion in 2026 and 2027—roughly the annual GDP of Singapore. But the Wall Street Journal has reported that American consumers spend only $12 billion a year on AI services. That’s roughly the GDP of Somalia. If you can grok the economic difference between Singapore and Somalia, you get a sense of the economic chasm between vision and reality in AI-Land. Some reports indicate that AI usage is actually declining at large companies that are still trying to figure out how large language models can save them money.
Every financial bubble has moments where, looking back, one thinks: How did any sentient person miss the signs? Today’s omens abound. Thinking Machines, an AI startup helmed by former Open AI executive Mira Murati, just raised the largest seed round in history: $2 billion in funding at a $10 billion valuation. The company has not released a product and has refused to tell investors what they’re even trying to build. “It was the most absurd pitch meeting,” one investor who met with Murati said. “She was like, ‘So we’re doing an AI company with the best AI people, but we can’t answer any questions.” Meanwhile, a recent analysis of stock market trends found that none of the typical rules for sensible investing can explain what’s going on with stock prices right now. Whereas equity prices have historically followed earnings fundamentals, today’s market is driven overwhelmingly by momentum, as retail investors pile into meme stocks and AI companies because they think everybody else is piling into meme stocks and AI companies.
Every economic bubble also has tell-tale signs of financial over-engineering, like the collateralized debt obligations and subprime mortgage-backed securities that blew up during the mid-2000s housing bubble. Ominously, AI appears to be entering its own phase of financial wizardry. As the Economist has pointed out, the AI hyperscalers—that is, the largest spenders on AI—are using accounting tricks to depress their reported infrastructure spending, which has the effect of inflating their profits. As the investor and author Paul Kedrosky told me on my podcast Plain English, the big AI firms are also shifting huge amounts of AI spending off their books into SPVs, or special purpose vehicles, that disguise the cost of the AI build-out.
My interview with Kedrosky received the most enthusiastic and complimentary feedback of any show I’ve done in a while. His level of insight-per-minute was off the charts, touching on:
How AI capital expenditures break down
Why the AI build-out is different from past infrastructure projects, like the railroad and dot-com build-outs
How AI spending is creating a black hole of capital that’s sucking resources away from other parts of the economy
... continue reading