The “AI Railroad” Idea

The most interesting idea in Porter Stansberry’s “Artificial Illusion” presentation is not a stock pick. It is a frame. Stansberry argues that the people getting rich from AI will not be the ones building the models, but the ones who own the rails the models run on. He calls these companies the “AI Railroad,” and the analogy is to the railroad barons who made fortunes hauling other people’s gold while the miners went broke digging for it.

It is a genuinely useful mental model, and it is also where the pitch quietly changes the subject. Once you accept the railroad frame, the next step is to point at some profitable, established businesses and call them the rails. That is where fintech enters, and where the distinction between infrastructure and AI beneficiary starts to blur.

Why Fintech Fits the Metaphor

Payment companies fit the railroad metaphor better than almost anything else in the market. Payment rails are, literally, infrastructure. Every online merchant needs them. They carry transaction volume regardless of what is being bought. A processor earns a fee on every payment that moves across its rails, which is about as close to a tollbooth as a software business gets.

That is why Stansberry’s pick of PayPal lands as the centerpiece of the pitch. PayPal, through its Braintree processor, does own real payment infrastructure. We unpack that in our Braintree explainer. The rails are real. The question is whether owning rails makes a company an AI beneficiary.

Where the Metaphor Breaks Down

Here is where the frame starts to strain. A railroad’s value comes from the fact that it is hard to build a competing track to the same destination. Payment processing does not have that barrier. Competitors can, and do, build competing rails. Adyen, Marqeta, Block, and Stripe all process payments, and merchants can switch between them or use several at once. The infrastructure is real, but it is not scarce the way a physical railroad is.

The bigger strain is on the “AI beneficiary” half of the label. Payment companies do use machine learning, mostly for fraud detection and routing. But that machine-learning use is a cost of doing business, not a source of pricing power. Every processor has it. It does not make a fintech stock go up the way an actual earnings catalyst does.

This is the same thread we pull in our look at what “AI” actually means in these promos. The label is doing more work than the technology.

Fintech Is Not an AI Bet

The honest way to think about it is that fintech stocks are a value proposition, not an AI proposition. The three names Stansberry teased, PayPal, Ulta Beauty, and Deere, are all real, profitable businesses. Two of them are not even fintech. Ulta is a retailer. Deere is a farm equipment maker. The only thing they share is that they are established, capital-efficient, and, in some loose sense, touched by machine learning.

That is a fine thing to buy. Buying profitable businesses at reasonable prices works over time. But it is not an AI call. The returns over the three years since the tease tell the story: the one pick that actually delivered, Deere, won on farming fundamentals, not on AI. We track that in the Deere scorecard. The AI wrapper added color, not return.

The One Real Lesson in the Metaphor

For all the places the “AI Railroad” frame breaks down, it does contain one transferable idea worth keeping. When a boom is underway, the surest money is often in the picks-and-shovels layer rather than the speculative frontier. That instinct, buy the boring supplier instead of the exciting miner, has a long and respectable history.

The error is not the instinct. It is the identification. A railroad is a natural monopoly with a moat that regulation and geography protect. A payment processor is a competitive software business where the moat has to be rebuilt every few years. Calling the second thing by the first thing’s name flatters it without improving it.

So the lesson survives, but in a humbler form. Before you buy “the rails of the next boom,” ask whether the company you are looking at actually has a rail-like moat, or whether it just has a plausible claim to one. Payment companies, for all their genuine infrastructure, are mostly the second thing.

The Bottom Line

The “AI Railroad” is a useful way to remind yourself that the infrastructure around a boom can be a better bet than the boom itself. That part of Stansberry’s thinking is worth keeping.

The mistake is the slide from “infrastructure” to “AI beneficiary.” Owning payment rails makes you an infrastructure company. It does not make you an AI company. Fintech stocks should be judged on transaction volume, pricing power, and valuation, not on how convincingly they can be described as AI. The railroad analogy is good; the AI label attached to it is mostly marketing.

Ready to see the research? Click here to access Porter Stansberry’s report.

NewsletterVetter is an independent publication. We receive compensation from some of the services we review through affiliate links. Nothing on this site is investment advice. Always do your own research.