AI Spending Is Now Bigger Than the Dotcom Bubble — and Cracks Are Showing
We've Been Here Before. Sort Of.
If you were old enough to have a in 1999, you remember the dotcom bubble — that era when investors poured staggering sums of money into internet companies because the internet was clearly going to change everything. They were right about the internet. They were very wrong about the valuations. Trillions of dollars evaporated when the bubble burst in 2000.
Now, a leading global bank is putting a number on something that's been nagging at economists and market watchers for a while: the current wave of AI investment has officially grown larger than the dotcom boom at its peak. Not just similar in spirit — larger in actual scale. And while that's a testament to how seriously the world is taking artificial intelligence, it's also the kind of milestone that makes careful analysts reach for the antacids.
What's Actually Happening
Over the past two years, the money flowing into AI infrastructure — the data centers, the chips, the software platforms, the cloud computing capacity required to train and run AI systems — has been extraordinary. We're talking about the biggest technology companies in the world committing hundreds of billions of dollars, often before they have a clear picture of exactly how those investments will pay off.
That last part is the key tension. During the dotcom era, the logic was: the internet is real, it's growing, so pour money in and figure out the business model later. A lot of those companies never found one. The bank's warning isn't that AI is fake — it's clearly not — but that the *spending* has raced so far ahead of *proven returns* that the gap is starting to look uncomfortable.
The phrase economists use for this is a "build-out ahead of monetization" problem. In plain English: companies are spending enormous amounts of money building the tools and infrastructure for AI before they've fully demonstrated they can make that money back. When the dotcom bubble burst, it wasn't because the internet failed — it was because the economics didn't work the way investors assumed they would, at least not on that timeline.
Where the Cracks Are Appearing
The bank's note flags that early signs of strain are becoming visible. Not a crash — not yet, and possibly not at all — but the kinds of signals that warrant attention. Some AI-focused companies are starting to face harder questions from investors about when exactly all this infrastructure spending translates into real profit. The market has been extraordinarily patient with the "trust us, the returns are coming" narrative, but patience has limits.
There's also a concentration problem. A huge portion of AI spending is happening inside just a handful of massive companies — the kind of firms that can absorb losses for years if they need to. But the broader ecosystem of smaller AI startups and vendors depends on that spending continuing. If the big players start pulling back, even slightly, the ripple effects could be significant.
At the same time, some of the early productivity gains from AI that were supposed to justify all this investment have been slower to materialize in measurable economic data than the hype suggested. That's not fatal — transformative technologies often take longer than expected to show up in the statistics — but it does mean the timeline for "AI pays for itself" is getting pushed out.
Why This Matters to You
You don't need to own a single tech stock for this to affect your financial life. If you have a retirement account — a , an IRA, a pension — there's a very good chance it has significant exposure to the companies at the center of this AI spending boom. The biggest AI investors are also some of the largest components of major stock market indexes, which means index funds, the default choice for millions of everyday investors, are heavily tied to how this plays out.
The bank's warning isn't a call to panic or to sell everything. It's more like a friend who works in finance pulling you aside and saying: *enjoy the party, but maybe know where the is.* The AI boom could absolutely justify its own hype over time. But the gap between what's being spent and what's being earned is wider than it's ever been, and that gap is something worth watching — even if your day job has nothing to do with servers or semiconductors.