Is the AI boom a bubble waiting to pop, or are we watching the early stages of something much bigger? Daniel Newman, CEO of Futurum and a well-known tech industry analyst, came down firmly on the "bigger" side during a recent conversation with Marley Kayden on Schwab Network.

The Supercycle Argument

Newman's take is straightforward: we're not in a bubble. We're in a "multi-decade technological supercycle." Sure, that sounds like something someone would say right before a bubble bursts, but Newman backs it up with specifics.

He admits there are bubble-like symptoms floating around. Some companies mention "AI" approximately 47 times per earnings call despite generating zero actual AI revenue. Others are spending enormous sums on AI initiatives without much to show for it, yet their valuations have soared anyway. These are legitimate concerns.

But here's Newman's key distinction: the core infrastructure companies—Nvidia (NVDA), Google (GOOGL), Microsoft (MSFT)—are driving real, measurable revenues. These aren't speculative plays. They're building the actual plumbing of the AI economy, and people are paying good money for it.

When Does AI Actually Make Money?

Newman highlighted that right now, we're in the infrastructure build-out phase. The big question is whether all this capital expenditure will eventually generate returns. Much of the current revenue is flowing to companies like OpenAI and Anthropic, but Newman believes the real monetization opportunity lies in enterprise applications, particularly what he calls "agentic use cases"—AI systems that can actually do things autonomously.

According to Newman, 2026 will be the "real year of proof" for the AI trade. That's when we'll see whether the massive CapEx investments translate into actual enterprise revenue and economic growth, or whether companies have been building extremely expensive infrastructure that nobody ends up using at scale.

The Real Risk: What If Nobody Shows Up?

Newman isn't saying everything is perfect. He pointed to a specific risk that could turn this supercycle into a regular old bubble: supply constraints flipping into a supply glut. "Where the bubble really could exist is if all of a sudden we had a glut, the demand for utilization goes down, and the use cases don't take off," he explained.

In other words, if companies build all this AI capacity but enterprises don't actually adopt AI tools at the expected rate, we'd have a problem. That's the nightmare scenario: tons of expensive infrastructure sitting idle because the killer apps never materialized.

The Broader Debate

Newman's perspective lands in the middle of a heated Wall Street debate. Google DeepMind CEO Demis Hassabis recently suggested some AI startups are overvalued and a market correction might be coming. Meanwhile, Goldman Sachs Asset Management has argued the AI buildout is structurally sound because it's funded primarily by corporate cash flows rather than risky debt.

The question of whether we're in an AI bubble remains one of the most pressing concerns for investors trying to figure out where to put their money. Newman's voice carries weight here—he's been analyzing the tech industry long enough to have seen actual bubbles pop. His view that the infrastructure layer is solid, even if some frothy speculation exists around the edges, offers a framework for thinking about where the real risks lie.

Whether he's right will become clearer in 2026, when the proof, as Newman put it, will be in the revenue pudding.