Remember Michael Burry? The guy who saw the 2008 housing crash coming while everyone else was still doing the Macarena at mortgage parties? Well, he’s back with another “uh oh” moment, and this time he’s got his sights set on the AI boom.
Here’s the deal: Burry thinks Big Tech is pulling some classic accounting shenanigans to make their AI profits look way shinier than they actually are. And honestly? It’s kind of genius in that “please don’t get caught” sort of way.
The Accounting Magic Trick
So here’s what’s allegedly happening. You know those fancy AI servers that cost more than your house? Companies like Meta, Oracle, and Google used to say “Yeah, these things will be obsolete in 3 years, so let’s expense them accordingly.” But now? They’re like “Actually, these babies will last 5-6 years easy!”
Why does this matter? Well, if you spread that $90 million server cost over 6 years instead of 3, you’re only hitting your books with $15 million per year instead of $30 million. Boom – instant profit boost! It’s like saying your iPhone will last 10 years instead of 3. Technically possible, but… come on.
Burry’s math suggests this little trick could be inflating Big Tech earnings by a whopping $176 billion through 2028. That’s not pocket change – that’s “buy a small country” money.
But Wait, There’s a Plot Twist
Here’s where it gets interesting. Maybe – just maybe – these companies aren’t lying. The AI gold rush is so intense that even “old” chips from 2021 are still chugging along doing useful work. When you’re desperate for compute power, nothing goes to waste. It’s like keeping your 2015 MacBook because hey, it still runs Netflix.
Plus, data centers have gotten way better at keeping hardware alive longer. Better cooling, smarter software, less wear and tear. So maybe extending those depreciation schedules isn’t fraud – it’s just reality.
The “Too Big to Fail” Factor
Even if Burry’s right about the sketchy accounting, there’s one massive difference between now and 2008: the government actually wants AI to succeed. This isn’t some random housing bubble – it’s treated like national security infrastructure.
When SoftBank’s CEO dumps $5.8 billion of Nvidia stock just to plow it into OpenAI and a $500 billion AI project, you know the smart money isn’t running away – it’s doubling down on the next phase.
The Bottom Line
Look, Burry’s track record speaks for itself. When he waves red flags, smart people pay attention. But this AI boom has something the dot-com bubble didn’t: ChatGPT hit a million users in five days and now serves 300+ million weekly. That’s not hype – that’s adoption at warp speed.
So what’s an investor to do? Keep your eyes open, watch those earnings quality metrics, and remember that even legendary investors can be wrong. The next few years will tell us whether Burry’s crystal ball is still working, or if this time really is different.
Either way, it’s going to be one hell of a ride.