Here’s a plot twist that keeps repeating in tech: the people who build the infrastructure go broke, and the people who use it get filthy rich.
Back in 1898, Campbell’s Soup hitched a ride on railroad tracks laid by the Philadelphia & Reading Railroad. Problem? The P&R was already bankrupt. But Campbell’s? They rode those rails straight to becoming a household name worth billions. The railroad company paid for the infrastructure. Campbell’s just showed up and made money.
This isn’t ancient history. It’s the playbook for every tech boom ever.
In the late 1990s, investors threw trillions at building the internet’s “rails”—fiber optics, telecom networks, all that unglamorous infrastructure. Cisco Systems became the poster child, then lost 85% of its value when the dot-com bubble burst. Brutal.
But the companies that *used* that internet infrastructure? Amazon went up over 100,000%. Google went up tens of thousands of percent. The builders got wrecked. The appliers got rich.
Now here’s where it gets interesting: those same internet winners—Alphabet, Amazon, Meta, Microsoft—are now spending aggressively to build AI infrastructure. They’re the new railroad companies. And if history is any guide, they’re about to become the new cautionary tale.
The real money in AI won’t go to the companies building the infrastructure. It’ll go to the companies *using* it to cut costs, boost productivity, and quietly expand margins in boring, established businesses.
Think about PayPal. Not exactly sexy. But it’s embedding AI everywhere—fraud prevention, transaction approvals, customer service, loan underwriting, even internal stuff like marketing and compliance. The result? Revenue per employee jumped over 50% since 2022. That’s not theoretical. That’s real efficiency gains happening right now.
PayPal’s got 435 million users already. It’s not building the AI rails. It’s using them to make its existing business run better. And with AI influencing more commerce every day, PayPal’s positioned to benefit without risking its balance sheet to construct the infrastructure.
Here’s the thing: the railroad companies didn’t fail because their idea was wrong. They failed because they built too early, spent too much, and couldn’t capture the value they created. The real fortunes went to the second movers—the ones who showed up after the infrastructure was already there.
We’re watching that exact shift happen with AI right now. The hyperscalers are laying the rails. But the appliers—the companies quietly using AI to get smarter, faster, and more profitable—those are the ones positioned to capture the biggest gains.
You don’t want the P&R companies. You want the Campbell’s.
The builders will struggle. The appliers will get rich. That’s not a prediction. That’s history repeating itself.
