The new oil? Inside the effort to turn AI computing power into a tradeable commodity

The push to commodify AI computing power marks a pivotal shift in how the digital economy manages risk, one that could reshape the infrastructure of innovation itself. Just as futures markets once tamed the volatility of physical commodities like oil and wheat, this emerging sector seeks to standardize and trade the most intangible yet essential resource of the 21st century: raw computational capacity. At its core, the effort reflects the growing recognition that AI’s true bottleneck is no longer algorithms or data, but the finite and expensive infrastructure required to train and deploy them. By creating liquid markets where companies can lock in access to GPUs and cloud compute, the startup at the center of this effort isn’t just selling a financial instrument—it’s attempting to impose stability on a market currently characterized by wild price swings, long waitlists, and opaque allocations. If successful, it would do for AI what derivatives did for global energy: decouple access from scarcity, turning a constrained resource into a predictable utility. This isn’t the first time markets have tried to monetize scarcity in tech. During the dot-com bubble, bandwidth futures briefly emerged as a speculative play, only to collapse when the infrastructure caught up with demand. The difference now is scale—and stakes. AI’s computational hunger dwarfs even the most voracious pre-internet industries. A single large language model can consume thousands of GPU-hours, and the global race to build them has already strained supply chains, triggering price surges that have priced out smaller researchers and startups. A futures market could democratize access, but it could also exacerbate inequities if only well-capitalized firms can afford to hedge against future shortages. Regulators, too, face uncharted territory: How do you prevent manipulation in a market where the underlying asset is itself a black box—shaped by rapidly evolving hardware, geopolitical chip restrictions, and the whims of semiconductor giants? The most pressing question is whether such a market can mature before the next AI boom hits. If compute becomes as tradable as crude oil, will we see AI hedge funds treating neural networks like trading algorithms? Or will the market prove too volatile, too fragmented, or too prone to gaming? The experiment begins now—and its outcome may determine who gets to shape the next era of artificial intelligence.