AI May Be the New Internet
But the Profits Haven't Caught Up Yet
Artificial intelligence is the story everyone’s chasing — but investors are discovering that technological revolutions and profit revolutions rarely move in lockstep. In the 1990s, the Internet transformed communication long before it transformed earnings. Today, AI is following the same script: a surge of innovation, a wave of speculative enthusiasm, and a widening gap between adoption and monetization.
This week’s Rational Edge essay explores why markets repeatedly overestimate near-term gains while underestimating long-term compounding. The pattern isn’t new — it’s behavioral. By studying AI through the same lens that once applied to the Internet, investors can learn how narratives inflate faster than cash flows — and how disciplined patience ultimately captures the real edge.
The Hype Cycle Returns
Every generation rediscovers the thrill of a new technology and the impatience that comes with it. Artificial intelligence, in 2025, sits exactly where the Internet did in 1998 — transformative in promise, thin in profits, and rich in narrative.
Startups describe “AI integration” as a business model. Corporations reference AI in every earnings call. Venture capital pours in, and headlines suggest an unstoppable wave of productivity. Yet, if we examine fundamentals, most AI-related companies still generate modest revenue, high expenses, and uncertain paths to durable margins.
Echoes of the Dot-Com Era
The comparison isn’t meant as cynicism but as context. The Internet revolution changed the world — but it took almost a decade before profits justified valuations. Between 1999 and 2002, the NASDAQ lost 78 percent of its value, even as broadband and e-commerce adoption accelerated. The innovation succeeded; the timing of profits didn’t.
AI today exhibits similar dissonance. The technology is advancing faster than the business models built on it. Large language models consume immense computational power, training data, and energy. Cloud infrastructure providers benefit first; downstream users struggle to convert productivity gains into cash flow. It’s the classic sequence of innovation: infrastructure profits before application profits.
The Psychology of Premature Extrapolation
Why do markets repeat this pattern? Because investors don’t value innovation — they value expectations. When possibility expands faster than evidence, imagination fills the gap. Behaviorally, this is called *premature extrapolation* — our tendency to project linear success from early breakthroughs. We saw it in dot-com valuations, cryptocurrencies, and now AI.
The narrative feels irresistible because it has truth at its core. AI will reshape every industry. But between invention and integration lies the slow terrain of business execution — process change, cost reduction, pricing power, and adoption curves. Investors love parabolas; real profits arrive as sigmoid curves.
Where the Money Actually Flows
In the near term, the economic beneficiaries of AI are not necessarily the companies that use AI, but those that enable it. Just as Cisco, Oracle, and Intel captured the first Internet profits, today’s AI winners are chip manufacturers, data-center operators, and cloud platforms. Infrastructure monetizes before application innovation.
That doesn’t mean AI-driven software firms won’t eventually succeed — it means the cash flow curve is delayed. Markets often discount future profits too quickly, creating a feedback loop of rising valuation and future disappointment. Investors who confuse technological adoption with profit timing become impatient when reality lags expectation.
Behavioral Blind Spots
Behavioral finance explains why intelligence alone isn’t an edge. The bias of recency — believing today’s trend is permanent — drives much of the AI mania. So does herding: the comfort of consensus that feels like truth. Investors see charts moving up and rationalize backward. Fear of missing out becomes a story about “not being left behind.”
At The Rational Edge, we frame FOMO not as greed but as anxiety misapplied to finance. It’s a signal that patience is being undervalued. When narratives expand faster than earnings, the calm investor doesn’t withdraw — they wait. They let time do what emotion cannot: reveal the real winners.
From Story to System
AI is the most important technological inflection point since the Internet, but the investment lesson isn’t new. Narratives ignite attention; profits require patience. The edge belongs to those who separate signal from story, novelty from durability.
The goal isn’t to avoid innovation — it’s to time exposure rationally. Study cash flows, not headlines. Follow the capital-expenditure chain before the consumer application. And remember that valuation and innovation operate on different clocks.
The Rational Takeaway
AI will reshape the world in ways we can barely measure — but profitability will arrive in chapters, not bursts. Investors who expect a linear translation from hype to returns will relive the lessons of 2000. Those who cultivate patience and clarity will discover that discipline, not prediction, is the true edge in an age of artificial intelligence.