UpTrajectory Review

Giuseppe Cicatelli's Entrepreneur piece makes a single, well-worn argument in its opening line: the smarter investment in AI is the infrastructure layer—the 'picks and shovels'—rather than the consumer-facing applications built on top of it. The full text isn't available to us, only that thesis statement, but the framing is instantly recognizable to anyone who has tracked technology cycles. It is the same logic that made Cisco, Intel, and later AWS and Nvidia the durable winners of the internet and cloud eras, while thousands of app-layer companies rose and fell on top of their platforms. Cicatelli is essentially saying: don't pick the winner of the AI race, sell the equipment every racer needs.

For a small-business operator, this isn't abstract portfolio theory—it maps directly onto procurement and strategy decisions happening right now. Every vendor pitching you an 'AI-powered' CRM, chatbot, or analytics tool is making an app-layer bet, and many of those bets will not survive the next eighteen months. The infrastructure providers—cloud compute, model APIs, data-pipeline tools, and the chip and networking suppliers behind them—have fundamentally different economics: recurring, usage-based revenue tied to volume rather than to any single product's popularity. If you're building workflows around an AI tool, the stability of what's underneath it matters more than the feature list on top.

The thesis is sound, but it deserves some pushback. 'Picks and shovels' is a comfortable analogy that can flatten real risk: infrastructure plays are capital-intensive, cyclical, and already heavily bid up by exactly the smart money Cicatelli is describing. The gap between a good sector and a good entry price is enormous. There's also a timing question the headline glosses over—by the time a trade is consensus enough to anchor an Entrepreneur column, much of the easy alpha has often been extracted. We're skeptical of any framing that treats infrastructure as a risk-free proxy; the history of railroad bonds and fiber-optic cable says otherwise.

The second-order effect worth watching is consolidation pressure on the app layer. If capital and margins concentrate upstream, the app companies competing for your subscription dollars will face a squeeze: they'll either get acquired, pivot to vertical niches where they own a workflow end-to-end, or fold. That has practical consequences—data portability, contract terms, and the risk that a tool you rely on gets sunset after an acquisition. It also means the talent and feature innovation that might have differentiated apps will increasingly migrate to the infrastructure companies themselves, which are already moving up the stack.

What to do with this: audit the AI tools you're paying for and ask what's actually underneath them. Who owns the model, the compute, and the data pipeline? A tool built on a major cloud provider's infrastructure with a clear API and export path carries different risk than a thin wrapper on a single third-party model. If you're making longer-term technology bets for your business, favor vendors whose economics don't depend on winning a crowded app race. And if you're an investor reading Cicatelli's argument as a stock tip, treat it as a starting point for valuation work, not a conclusion—the right layer and the right price are different questions.

“Here's why I'm betting on the picks and shovels of AI, not the apps.” — Entrepreneur

Takeaway: Before subscribing to any AI tool, check what infrastructure it runs on and whether your data can move if the app-layer company doesn't survive.