Remember when everyone made bank on Nvidia? Yeah, those days are basically done. But here’s the thing: the AI boom isn’t over. The *leaders* are just getting tired.
Think about it like this: the Magnificent Seven—Nvidia, Meta, Microsoft, Alphabet, Apple, Amazon, and Tesla—were the first to the dance. They built the tech, they got the glory, and early investors made absolutely ridiculous returns. Nvidia alone is up over 1,000% since ChatGPT launched. That’s the kind of move that makes people quit their jobs.
But now? These companies are burning through cash like it’s going out of style.
Meta just raised its 2026 capex forecast to $125-145 billion. Alphabet? $180-190 billion. Amazon, Microsoft, Meta, and Alphabet combined spent nearly $300 billion on capital expenditure last year. This year they’re doubling that to $635 billion. That’s not an investment strategy—that’s a prisoner’s dilemma where everyone’s overinvesting simultaneously because they’re terrified of falling behind.
Here’s where it gets interesting: OpenAI, the poster child of the AI revolution, lost $8 billion last year on $12 billion in revenue. This year they’re expecting $17 billion in losses. Next year? $35 billion. These aren’t the numbers of a company about to print money.
The real problem is that AI is becoming a *cost center*, not a growth driver. Even if revenues keep climbing, margins compress, expectations reset, and valuations shrink. The Mag 7 aren’t going anywhere, but their best days? Probably in the rearview mirror.
**So where’s the second chance?**
Here’s what most investors get wrong: they assume the companies that *build* new technology automatically capture the biggest returns. Spoiler alert: they usually don’t.
The real money in the next phase of AI won’t come from the hyperscalers burning billions on infrastructure. It’ll come from the companies *using* that technology—the ones hiding in plain sight, not typically thought of as AI companies at all.
These are firms that are becoming AI companies quickly, effectively, and without the nosebleed valuations of the Mag 7. They’re getting in at ordinary, even *low* valuations while positioned to grow rapidly. It’s like getting in on Nvidia years ago, except you’re not paying Nvidia prices.
The first wave of AI gains went to the pioneers. The second wave? That’s going to the smart money that figures out how to *use* what the pioneers built.
If you watched others make a fortune on AI and thought you missed it, think again. The real opportunity is just starting. And unlike the Mag 7’s bloated capex budgets, this one doesn’t require betting the farm.
