Intelligence as a commodity

For most of history, intelligence was rare. It lived in people’s heads : scientists, analysts, writers, and workers who thought for a living. But today, with LLMs, intelligence has become something you can buy, a service available through an API. You pay per question, per reasoning step, or per thousand tokens.

Just like electricity or cloud storage, intelligence is turning into a commodity.

When I say “intelligence as a commodity,” I mean that the ability to think, reason, and write is now a measurable, tradable product. Benchmarks compare models using both accuracy and cost: how many dollars it takes to complete a task.

Cheaper models are like budget airlines: basic and good enough for many jobs. More advanced models (with deeper reasoning or “thinking longer”) are like business class: pricier and more capable. Over time, both get cheaper as technology improves.

This creates a fascinating dynamic: as the cost per task drops, the number of tasks we’re willing to automate explodes. We start using AI for things that once seemed too small or too expensive : from summarizing every email to helping plan our week.

Behind every AI query lies a massive physical infrastructure: data centers packed with GPUs, miles of fiber-optic cable, and endless cooling systems humming in the dark.

As intelligence becomes cheaper and more widespread, our appetite for it grows and so does the demand for the hardware that powers it. Every reasoning step, every generated sentence burns energy, silicon, and capital.

The irony is that while “intelligence” feels weightless, its backbone is incredibly heavy: steel, copper, electricity, and land. Data centers are becoming the new factories ; the industrial machines of the knowledge age.

And just like any industrial revolution, the early years are marked by massive investment and intense competition. The goal: to make intelligence cheaper, faster, and more accessible until it’s as ubiquitous as Wi-Fi.

What happens when intelligence becomes nearly free?

• Every product and service can have a built-in brain.
• Every decision, personal or corporate, can be assisted by reasoning algorithms.
• The bottleneck shifts from access to imagination: what new things can we build once intelligence itself is abundant?

We may end up living in a world where intelligence is like electricity, it's barely visible, is everywhere, and indispensable. The winners won’t be those who own the most intelligence, but those who know what to do with it.