The U.S. market has never been more focused on one single bet—but AI’s payoff will likely take decades

America’s AI boom is starting to look less like a tech cycle and more like a national wager. Fortune says AI-related capital spending has climbed to about 2% of U.S. GDP from near zero five years ago, and the money is piling into a very small set of companies and assets.

The scale is what makes this feel different. In January, the Federal Reserve Bank of St. Louis said investment tied to information processing equipment, software, research and development, and data centers accounted for 39% of U.S. economic growth in the first three quarters of 2025. That is more than the 28% share those categories reached at the dot-com peak in 2000. Pantheon Macroeconomics also says that without AI spending, U.S. corporate investment in equipment would have been negative last year.

The stock market is just as narrowed. Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia, and Tesla now make up a combined $25 trillion in market value, or 35% of the S&P 500. That is the highest concentration in a handful of stocks since just before the crash of 1929. The credit market is leaning the same way, with AI-related companies’ share of investment-grade bond issuance rising from 2% in 2023 to 10% in 2026, according to Federal Reserve Bank of Kansas City research.

Then there is venture capital. AI took 81% of a record $300 billion in global venture funding in the first quarter of 2026, and four companies — OpenAI, Anthropic, xAI, and Waymo — drew two-thirds of it. At the same time, the top 20% of U.S. earners account for about 60% of consumer spending, according to Moody’s Analytics, and their wealth is tied to the AI stock rally. That creates a loop: spending lifts growth, rising valuations enrich the richest households, and that extra spending feeds back into the economy.

The problem is that the loop is also fragile. The Bank for International Settlements warned in its 2026 Annual Economic Report that AI exuberance could end in a “lengthy investment bust,” and it flagged circular financing as a core vulnerability. More than half of investment in AI firms comes from other AI firms, with the money often used to buy the backers’ own products. That is not exactly the kind of healthy demand curve economists put on posters.

The real assumption behind the whole thing is that by 2027, infrastructure spending will give way to broad productivity gains across the rest of the economy. The piece argues that is likely too optimistic. AI may eventually deliver, but only if hospitals, schools, logistics companies, retailers, and everyone else can actually adopt it, not just buy the chips and build the data centers. That sounds less like a sprint and more like a very expensive waiting room.