Let's talk about the part of AI nobody wants on a slide.

The International Energy Agency estimated that data centers consumed around 415 terawatt-hours of electricity in 2024, roughly one and a half percent of the world's total. Their projection: about 945 terawatt-hours by 2030, more than doubling, with AI the main driver.

That's about the electricity use of Japan, on its way to more.

What's the pattern?

Here's what history says. Every era runs on an energy source, and whoever controls it writes the rules.

Coal built the British Empire. Oil built the American century. In Radical Next I write about the Kardashev scale, the physicist Nikolai Kardashev's idea that a civilization's advancement can be measured by how much energy it can harness. He was thinking about stars. But the logic works at the scale of nations and companies too.

The next era's energy question isn't just how much. It's where, how fast, and how clean. Chip wars are really energy wars. Data center sites are chosen by grid capacity before anything else. Countries with cheap, reliable power are becoming the new resource states, and countries without it are becoming customers.

And there's a second-order effect. When compute costs are dominated by electricity, the price of thinking becomes the price of power. That reshapes who gets to build, and where.

What should leaders do now?

So, three things.

Know your AI's energy footprint before your board asks. It's coming as a question and a cost line.

Treat power as strategy. If your operations depend on compute, your energy contracts are now part of your competitive position. Most companies leave that to facilities.

And watch the grid, not just the model. The bottleneck of the next five years isn't chips. It's transformers, transmission lines, and permits. The organizations moving early on those will have compute when others are waiting.

Every era has an energy question. This one arrived faster than the last, and it's already on your bill.

Do you know what your AI cost you in electricity last month?

Related: How do you lead through uncertainty without pretending to be certain?

Sources
  1. International Energy Agency, Energy and AI, April 2025.
  2. Sam Rad, Radical Next (2025), on energy and the Kardashev scale.

Sam Rad, The Change Futurist. Keynote speaker on change, transformation, resilience, and AI adoption. Author of Radical Next. Book a keynote