If you’ve lost sleep wondering whether you’re missing the biggest opportunity since the internet—or about to be the bag holder in a dot-com-style crash—you’re not alone. That paralyzing anxiety is the real fuel of today’s AI gold rush.
The AI boom isn’t a technology revolution. It’s a wealth transfer from the desperate to the prepared.
You’ve probably noticed every vendor suddenly claiming to be “AI-powered.” Your CRM, your accounting software, even your coffee machine. It’s not because they’ve all discovered magic. It’s because the infrastructure sellers—Nvidia, Microsoft, Amazon—have brilliantly positioned themselves as the only path to salvation. They’re charging a “compute tax” on every company’s deepest fear: missing the next wave.
History has a brutal lesson for this moment. The railroad boom of the 1840s? The telegraph craze in the 1860s? The internet bubble of the late 1990s? In each case, the underlying technology was genuinely world-changing. But the early infrastructure investors overpaid by an order of magnitude. They bought tracks when they should have bought the goods moving on them.
The biggest winners of the internet weren’t the fiber-optic cable layers. They were the application builders who didn’t arrive until the hangover was over.
Let me be direct: the current AI frenzy is less about intelligence and more about infrastructure providers extracting a tax on corporate FOMO. Every enterprise that rushes to buy GPUs and rent cloud compute is paying a premium for shiny shovels while the real gold lies hidden in the applications that will come later.
I saw this firsthand in 1999. My mentor at the time was a telecom analyst who warned that the fiber being laid under every street was real, but the companies laying it would go bankrupt. He was right. The actual value was captured a decade later by Netflix, Amazon Web Services, and Spotify—builders who used that infrastructure to create platforms no one had imagined.
Here’s the twist: the AI technology today is absolutely transformative. That’s what makes this paradox so painful. The revolution is real, yet the capital deployment is completely detached from near-term economic reality. You can be right about the future and still lose everything by buying in at the wrong stage.
Stop investing in picks and shovels. Start looking for the people who will build the cities.
The smart money right now isn’t chasing the latest GPU allocation. It’s studying where the real application-layer value will emerge—the companies that will use AI to solve specific, painful problems for businesses and consumers. Those second movers won’t arrive for 18 to 36 months. But when they do, they won’t be paying the compute tax. They’ll be collecting it.
So here’s my take: the AI infrastructure boom is a massive transfer of wealth from anxious investors to savvy hardware and cloud providers. The bubble will correct. The fundamentals will survive. And the biggest fortunes will be made by those who have the patience to wait for the real builders to show up.
You’re not crazy to feel torn. But now you know which side of history you want to be on.
FAQ
Q: Are you saying AI is a bubble and we should avoid it entirely?
A: No. I'm saying the technology is real, but the current infrastructure investment frenzy is detached from near-term reality. The value will shift to applications in 2-3 years. Avoid overpaying for commoditized compute; position for the application layer.
Q: What's the practical implication for my portfolio or business strategy?
A: Stop allocating capital to GPU farms, cloud credits, or generic AI infrastructure. Instead, fund or invest in narrow AI applications that solve specific, high-value problems—like automated contract analysis, medical diagnostics, or supply chain optimization.
Q: But what if this time is different? AI could be bigger than the internet.
A: Every boom thinks it's different. The internet was bigger than railroads, yet the infrastructure investors still got crushed. The pattern holds: the more transformative the tech, the bigger the over-investment in infrastructure, and the more dramatic the shakeout. History doesn't repeat exactly, but it rhymes.