I've been watching the Michael Burry debate heat up again, and honestly, it's getting harder to take his latest AI bubble thesis seriously. Let me break down why.

Burry's legendary status is undeniable—the guy made roughly $100 million personally and $700 million for investors by nailing the 2008 subprime crisis. That's the kind of track record that gets you a Christian Bale biopic. But here's the thing: his michael burry net worth and reputation don't automatically make him right about everything. Since that 2008 windfall, his performance has been... inconsistent at best. He's been repeatedly wrong on the markets, constantly calling doom that never materializes, and eventually shut down his hedge fund last year because he couldn't align with market reality.

Now he's claiming AI stocks are basically Dot Com 2.0. Let's examine his three main arguments because they're worth unpacking.

First, he argues that big tech firms like Meta, Microsoft, and Alphabet are cooking their books with depreciation schedules to pump earnings. Sure, GPUs have shorter lifespans than traditional servers, but here's what he's missing: most AI infrastructure actually has a 15-20 year useful life. Plus, older GPUs don't just vanish—they run inference tasks, so they retain value. The math doesn't support his fraud narrative.

Second, Burry warns that massive CAPEX spending will destroy cash flow. Except it's literally the opposite. Alphabet's operating cash flow jumped from under $100 billion to $164 billion in 2026. Margins are expanding across the board. And get this—companies scaling AI are reporting returns exceeding $3 for every $1 invested. The latest agentic AI wave is helping firms cut costs by 25% or more. That's not a sign of a broken model.

Third, he compares NVIDIA to Cisco in 2000, claiming both are overvalued. This comparison falls apart immediately. Cisco's P/E when it peaked in March 2000 exceeded 200. NVIDIA's current P/E sits at 47. Not even close.

The real market signal? H100 GPU rental prices have surged roughly 17% since mid-December, pointing to sustained scarcity and demand. That's bullish for infrastructure plays like Nebius Group, CoreWeave, and IREN. The energy constraint is real, which benefits companies like Bloom Energy that solve the power bottleneck.

Speaking of which, the options market is pricing in major moves. Deep-pocketed traders dropped a million-dollar bet on Bloom Energy call contracts, and one whale just put roughly $9 million on March $205 calls. Bloom shares jumped 8% on Monday while the Nasdaq stumbled, and the stock is forming a textbook weekly bull flag.

Look, Burry's contrarian legacy is solid, but the data on AI infrastructure is overwhelming. Between soaring H100 rental demand and the massive efficiency gains from agentic AI, the bull case remains intact. His michael burry net worth and track record don't change the fundamentals we're seeing in real time.