The money financing the AI boom may be cycling through a closed loop — the same handful of firms funding each other's spending, in what Digital Information World calls a "circular money machine." That phrase is doing a lot of work. And it's worth unpacking before the next quarterly earnings cycle convinces you the boom is self-sustaining.

The core question sounds almost childish: who's paying whom? In a healthy market, customers pay vendors, vendors pay suppliers, and profits flow outward. In today's AI economy, Digital Information World argues, the arrows bend back on themselves. Investors fund AI companies; those companies buy compute from cloud giants; the cloud giants' valuations rise on AI demand; the same investors reap the gains and reinvest. Around it goes.

What makes the money circular?

Picture the flow. A venture fund backs an AI startup. The startup spends most of its capital on GPUs and cloud credits — often from the very hyperscalers whose shares sit in the fund's portfolio. The hyperscaler books that as revenue. Revenue drives the stock. The stock returns money to the fund. Nobody outside the loop has to buy anything for the machine to keep spinning.

It's not illegal. It's not even unusual in young industries. Railroads did it in the 1870s. Telecoms did it before 2000. But circular revenue tends to inflate the apparent size of a market, because the same dollar gets counted several times as it travels the loop.

Why does the bubble question keep coming up?

Digital Information World frames the story as a boom "or bubble," and the hedge is the point. The distinction hinges on one variable: whether end customers — businesses and consumers outside the loop — eventually generate enough real demand to replace the internal spending.

If they do, the loop straightens into a pipeline. If they don't, the machine keeps running only as long as new capital keeps entering. That's the classic definition of a bubble, dressed in GPU-shaped clothing.

The piece doesn't declare the boom fake. It raises the accounting question an auditor would ask: follow the money, and count how many times it changes hands inside the same small circle of counterparties.

Who carries the risk if the loop breaks?

Concentration is the vulnerability. When a few investors, a few cloud providers and a few AI labs account for most of the spending, a single weak link — one slowing investment round, one hyperscaler cutting capex — transmits instantly through the chain. The circularity that amplifies gains on the way up amplifies losses on the way down.

There's a second-order risk, too. Circular spending can mask how little external revenue exists at any given moment. Markets price AI exposure on reported growth; reported growth, in a loop, partly reflects money the sector paid itself.

What would prove the boom is real?

The test is substitution. External demand — enterprises buying AI because it cuts costs or lifts output, not because they're invested in the ecosystem — has to become the dominant revenue source. Watch for that shift in the counterparty mix, not in headline growth rates, which the loop can inflate on its own.

Until then, the honest answer to "who's paying whom?" may be: largely, themselves. Digital Information World's analysis leaves the verdict open, but its framework gives investors a simple discipline — trace every dollar to a customer who can't sell you something back.