Google just dropped a number that made Wall Street's collective jaw hit the floor: $175 to $185 billion in AI infrastructure spending for 2026. To put that in perspective, analysts were expecting around $120 billion. That's not a rounding error—that's a 50% miss on expectations.
And here's the kicker: Google beat earnings. Revenue came in at $113.8 billion, crushing the $104.8 billion estimate. Earnings per share hit $2.82 versus the expected $2.61. By every traditional measure, this was a blowout quarter.
So why did the stock drop in after-hours trading? Because Wall Street is having a crisis of faith about AI spending.
The Bull Case: They Know Something You Don't
Look, Sundar Pichai isn't stupid. Neither are the folks running Microsoft, Amazon, or Meta. These companies have the most user data on the planet. They can see exactly how people are using AI tools, how much revenue it's generating, and what the trajectory looks like.
If the data showed AI was a dud, do you really think they'd keep doubling down quarter after quarter? Google's cloud division grew 47.5% year-over-year. That's not coming from people buying more Gmail storage.
The simple truth is this: companies with the best information are betting big. Either they're all wrong, or the market is pricing in too much short-term pain.
The Bear Case: This Is the New Metaverse
Remember when Mark Zuckerberg burned tens of billions on the metaverse? For years, he insisted it was the future. Wall Street rolled their eyes, and eventually he had to quietly scale it back.
The fear is that AI infrastructure spending is the new metaverse—a massive capital sink with uncertain returns. Spending $175 billion doesn't guarantee $175 billion in profit. It doesn't even guarantee $17 billion in profit.
And here's what should worry you: Google's free cash flow only grew 1% despite revenue being up 18%. All that extra money they're making? It's going right back into data centers, chips, and cooling systems.
