Been sitting with a thought recently.

Right now we’re in the golden moment of AI. The gold rush. The efficiency gains. The constant new features.

At the same time everyone’s worried about their job. Their relevance. Some are anxiously upskilling and experimenting to stay ahead.

But there’s a much bigger question sitting in plain sight that almost nobody’s talking about.

Where’s all the power going to come from?

AI isn’t magic. It’s computation on a massive scale, energy-intensive computation.

Data centers are already consuming vast amounts of power. And their numbers are growing fast.

So we’re talking about an exponential increase in power demand at precisely the moment we’re supposed to be reducing our energy consumption to tackle climate change.

So where’s this power going to come from?

More fossil fuels? Nope, that undermines every climate commitment we’ve made.

Renewable energy? We’re not scaling that fast enough to meet current demand, let alone the explosion coming from AI.

Nuclear? Maybe, but that takes decades to build and faces massive public resistance.

That means the current AI growth trajectory isn’t just ambitious. It’s unsustainable without a reimagining of global power infrastructure. Tech companies are racing ahead anyway.

The world’s data centers already consume roughly 1-2% of global electricity. That might not sound like much, but it’s equivalent to the entire power consumption of some countries.

And AI is pushing that number up dramatically. Training a single large language model (LLM) can use as much energy as several homes consume in a year. Now multiply that by the thousands of models being trained, the millions of queries being processed daily, with constant refinement and iteration.

The gap between AI’s power appetite and our ability to generate clean energy is widening, not closing.

What if data centers start competing with homes and hospitals for electricity? What if we have to choose between powering AI models or powering… cities?

It’s not a hypothetical future. In some regions, it’s already happening. Data centers are straining local power grids. Energy companies are struggling to meet demand.

I don’t have solutions. I’m not an energy policy expert. I’m not trying to add to the doom mongering on AI either.

I’m just someone watching this unfold and wondering why the question isn’t front and center in every conversation about AI’s future.

At some point, we’re going to hit a wall and be forced to choose. I say “we” but we all know I mean the people or companies in positions of power.

Slow down AI development? Not likely.

Serious environmental compromises? Probably. Trump will allow this, that’s for sure.

Power shortages? Most likely, and it will be everyday people who suffer.

Right now we’re in the golden moment of AI. It won’t (can’t) last, so it’s even more urgent for us to make the most of it. Anyone who delays may not get the same opportunity in 2-3 years.