This AI boom feels a little… 2008-ish

I’ve been riding the AI hype train just like everyone else. It’s impossible not to get excited about the breakthroughs happening almost daily. But lately, I’ve been getting this weird feeling of déjà vu, a little twitch that tells me something’s off beneath the surface.

Everyone’s talking about how AI is going to change the world, and they’re not wrong. But they were saying the same thing about the internet back in the late 90s, right before the dot-com bubble burst and wiped out trillions in shareholder value.

So, what’s going on now? We’re in the middle of an absolutely mind-blowing construction boom for AI data centers. The amount of money being thrown at this is staggering, and it’s become the secret engine keeping the U.S. economy afloat. Is this just another tech boom, or are we building a house of cards on a foundation of risky debt?

Let’s dig in.

⚙️ The Great AI Build-Out: A Firehose of Cash

First, you have to understand the sheer scale of this thing. AI, especially the advanced models we use today, is incredibly power-hungry. It takes a ton of computing power, or “compute,” to train these models, but the real cost monster is “inference”: the processing needed every single time you ask ChatGPT a question or generate an image.

As AI gets smarter, its need for compute is exploding. And a few tech giants have decided they are NOT sitting this one out.

Companies like Microsoft, Google, Meta, and Amazon are spending money like it’s going out of style. We’re talking a collective $100 billion+ on capital expenditures (capex) in a single quarter. For Microsoft and Meta, this spending is now more than a third of their total sales. That’s insane.

This isn’t just some line item on a balance sheet; it’s so massive it’s acting as a private-sector stimulus program.

According to some economists, this AI capex spending has contributed more to U.S. economic growth in the last two quarters than all of consumer spending combined.

Let that sink in. A handful of tech companies are literally propping up the economy by building these digital factories.

This reminds me of other huge infrastructure booms in history, like the railroad boom in the 1800s or the telecom boom in the 1990s. And you know what they both have in common? They ended in spectacular crashes.

✨ A Tale of Two Crashes: Why 2008 Was Way Scarier Than 2000

When people hear “tech crash,” they think of the dot-com bust in 2000. It was brutal, for sure. Fortunes were lost, Pets.com went belly-up, and a lot of investors got burned.

But here’s the key difference you absolutely need to understand: the dot-com bust was mostly an equity bubble. It was funded by stockholders and venture capitalists. When it crashed, shareholders lost their shirts, but the banking system was fine. The financial plumbing of the economy didn’t break. Everyone just walked away a little poorer and a lot more skeptical.

The 2008 crisis was a completely different beast. That was a debt-fueled bubble. The housing boom was built on a mountain of loans, specifically mortgages. Critically, these loans were held by banks. When the housing market collapsed, homeowners defaulted, and the banks that held that debt started to fail. This threatened to take down the entire global financial system. That’s a financial crisis.

Think of it this way:

• Equity Bust (2000): Your rich uncle invests in a flashy startup that fails. He loses his money. It sucks for him, but your savings account is safe.
• Debt Bust (2008): Your uncle borrows money from your bank to buy a house he can’t afford. He defaults. The bank loses its money, goes under, and suddenly your savings account is at risk.

So, the million-dollar question for the AI boom is simple: Who is funding it?

✍️ Following the Money: Where is it all Coming From?

When we look at who’s paying for all these shiny new data centers, we see a few different sources. Paul Kedrosky, a pundit who has been all over this story, breaks it down:

• Internal Cash: The big guys like Google and Microsoft are using their own massive cash reserves. This is the safest way to fund it. It’s their own money on the line.
• Stock & Bond Sales: Companies are also issuing new stock and selling bonds to the public. Again, this is pretty standard stuff. If things go south, bondholders and shareholders take the hit. It’s painful, but not necessarily world-ending.
• Venture Capital / Private Equity: Startups in the space, like CoreWeave, are getting huge checks from VCs and PE firms. This is the classic high-risk, high-reward model.

So far, this all sounds a lot like the dot-com era. But there’s another, murkier source of funding that’s growing at an alarming rate. And this is where my alarm bells start ringing.

🚨 The Shadow Banker: Everyone, Meet Private Credit

Have you heard of “private credit”? If not, you will. These are the new shadow bankers of the financial world, and they are at the heart of the AI boom.

Think of private credit funds as the debt version of private equity. They are firms that raise money from big investors (like pension funds and insurance companies), borrow even more money, and then lend it out directly to companies, often bypassing the public markets and regulated banks entirely.

Why is this scary? Three reasons:

1. Opacity: It’s a “private” market, meaning it’s a black box. It’s incredibly difficult to see who is lending to whom and on what terms. Regulators have a hard time keeping track of the risks building up inside.
2. Explosive Growth: This market has grown from a niche player to a multi-trillion-dollar giant in just a few years.
3. Concentration: They are pouring money into the AI data center boom. One of the biggest players, Blackstone, has a $100 billion war chest just for data centers.

So, these opaque, fast-growing funds are making massive, concentrated bets on a single sector. What could possibly go wrong?

⛓️ The Chain of Risk: How This Could Actually Break the System

Okay, so some weird investment funds are lending money. Who cares? Well, you should care. Because the next question is: where do they get their money from? And the answer is starting to look uncomfortably familiar.

A huge chunk of it comes from good old-fashioned banks.

A recent Federal Reserve note showed that bank lending to these private credit funds has skyrocketed. In 2013, it was 1% of banks’ loans to non-bank financial firms. Today, it’s 14%. Banks are handing these shadow lenders massive credit lines, which the funds then use to make their risky loans to data center developers.

Here’s the chain of contagion:

1. An AI company (let’s call it AI-Build-It) borrows billions from a Private Credit fund.
2. That Private Credit fund borrowed a lot of that money from a major, systemically important Bank.
3. The AI hype cools down. AI-Build-It can’t generate enough revenue to pay back its loans and defaults.
4. The Private Credit fund takes a massive loss and can’t pay back its loan to the Bank.
5. The Bank is now sitting on a huge loss, its balance sheet is damaged, and panic begins to spread.

And it gets worse. Who are the other big investors in private credit? Insurance companies.

A recent paper noted ominously that life insurers’ exposure to this kind of risky corporate debt now exceeds their exposure to subprime mortgage-backed securities back in 2007. Remember AIG? It was an insurance company that needed one of the biggest bailouts in 2008.

The real danger here is “correlation.” If all these private credit funds are lending to the same type of project (AI data centers), their fates are all tied together. If the AI hype train derails, it’s not just one car that goes off the tracks. It’s the whole train. That’s the “tail risk” the Fed is worried about.

✨ So, Are We Doomed?

Let’s be clear: this isn’t 2008… yet. But if you wait until it looks like 2008 to start worrying, you’re already too late. The classic ingredients for a financial crisis are starting to fall into place:

• ✅ A big, transformative story about why “this time is different” (The AI Revolution).
• ✅ A massive investment boom fueled by a rising tide of debt, all concentrated in one sector (Data Centers).
• ✅ An opaque, fast-growing, and lightly regulated corner of the financial system at the center of the action (Private Credit).
• ✅ Systemically important players, like banks and insurers, getting tangled up in the web.

Even Jamie Dimon, the CEO of JP Morgan Chase, is sounding the alarm, warning that private credit could trigger the next crisis. But at the same time, his own bank is jumping into the market.

It reminds me of what a former CEO of Citibank said after 2008: “As long as the music is playing, you’ve got to get up and dance.”

The music is definitely playing right now. The party is in full swing. But we need to keep a close eye on who’s going to be left without a chair when it stops.

More on This Topic

• The financial scale of the data center boom is immense. In 2023, the combined capital investments of Google, Microsoft, and Amazon in this sector surpassed the total capital spending of the entire U.S. oil and gas industry, highlighting a massive shift in economic investment.
• Energy demand is a primary concern. Projections indicate that by 2030, data centers could consume up to 9% of all electricity in the United States. This surge is forcing some utility companies to delay the retirement of fossil-fuel plants, complicating the transition to renewable energy.
• Industry experts are divided on whether this rapid growth constitutes a bubble. While some leaders warn of speculative overbuilding without firm customer commitments, others point to high pre-leasing rates for new facilities and power grid limitations as natural checks on runaway expansion.