Intelligence as a Utility Bill

We are witnessing the transformation of thinking into infrastructure.

For the entirety of human history, intelligence was bound to biology. If you needed strategy, analysis, or translation, you had to find a human brain. Today, that connection is permanently broken. Intelligence is becoming a utility bill—a metered, scalable flow of cognition that you plug into, pay for by the token, and scale up or down like electricity or water.

But as intelligence turns into raw infrastructure, we are forced to ask a question we have never had to answer before:

If cognition is becoming a utility bill, what is the bill actually for?

Cheap Intelligence, Unequal Power

Once intelligence becomes cheap, the real question is no longer who can think. It becomes who can direct that thinking, who can afford to scale it, and what they choose to aim it at.

This is where the formula of power comes in. Power has never been a function of intelligence alone. It has always been intelligence plus leverage.

In the agrarian era, leverage was land and physical labor. In the industrial era, it was capital, steam, and factories. In the digital era, it has been code and distribution networks. When you combine scalable, cheap cognition with modern leverage, the concentration of power doesn’t democratize—it accelerates.

AI turns basic thinking into a commodity. But because everyone gets the chatbot, the chatbot itself ceases to be a competitive advantage. The leverage shifts entirely to those who own the underlying assets: the proprietary datasets, the physical infrastructure, the capital, the distribution channels, and the sovereign authority.

Everyone gets the chatbot. Not everyone gets the factory.

“When intelligence becomes abundant, raw intelligence becomes less scarce. The scarce things become leverage, judgment, trust, and taste.”

The Tanzania Contrast

This is where the AI conversation becomes geographically and morally strange. Humanity does not have one shared future. It has multiple, vastly different futures happening at the exact same time.

I was recently in Tanzania, speaking to locals. In the Silicon Valley echo chamber I live in, the daily discourse is saturated with AGI timeline debates, token throughput, and whether humans are about to become obsolete. But on the ground there, nobody was talking about AI.

They were talking about how to have a normal life, where their kids could go to school, and how to secure a $200-a-month salary.

We were asking questions, learning about their country, trying to understand their history and the complex web of local cultures. Interestingly, some of the locals didn’t really know the details. Meanwhile, we were sitting there literally using ChatGPT and AI search tools to learn, validate, and parse their history in real-time. As foreigners, we ended up knowing more about their local history than the people living it, simply because we had the most advanced cognitive tools in the world in the palm of our hands.

The disparity of leverage was stark. And then, we saw the animals—the elephants and lions on the savannah. They didn’t give a fuck about AI. They were just there to get their day past and watch the grass grow.

The Priority Gap

That contrast is hard to sit with. In one part of the world, we are spending hundreds of billions of dollars building machines that write corporate emails, generate targeted advertising, optimize sales funnels, and summarize meetings. In another, basic physical survival remains unsolved.

The numbers make this priority gap brutal:

• The World Food Programme (WFP) reports that 363 million people are at risk of acute hunger in 2026.
• WHO and UNICEF estimate that 2.1 billion people still lack safely managed drinking water.
• Meanwhile, Stanford’s AI Index reported $252.3 billion in corporate AI investment in 2024 alone.
• The IEA projects data-center electricity use could double to about 945 TWh by 2030.

Too much of this intelligence is being pointed at small commercial optimizations before it is pointed at large human needs.

This is not a naive argument that AI is bad. It is an argument that the real test of our era is not AI capability—it is human direction. Power without purpose is just acceleration. And acceleration without humanity is not progress. It is just speed.

What is it for?

Ultimately, people do not want tokens. They do not want models, and they do not want productivity dashboards. They want:

• Less fear.
• Safe children.
• Food on the table.
• Clean water from the tap.
• Work that gives dignity.
• Health, meaning, and enough agency to breathe.

If metered intelligence cannot move us closer to solving these ancient, physical, human problems, then we have built the greatest cognitive engines in history and pointed them at the smallest possible goals.

The tragedy of AI will not be that it destroys us.

The tragedy will be that it reveals our priorities, and we fail the test.

Maybe the future of humanity will not be decided by whether AI becomes smarter than us, but whether we become wise enough to use it for anything that actually matters.