We’re often told to choose: grow the economy or protect the climate. Power the AI build-out or hit net zero. Every version of the choice assumes these ideas are add odds with one another.

I reject this premise. The future worth building is one where every generation can expect more than the last and the atmosphere gets healthier at the same time. Most people want that future too. So the real question is why does it feel so far away?

Part of the answer is policy. In many markets, the rules themselves hold progress back. Duplicative review processes make new clean energy siting costly. Permitting makes grid buildouts lengthy. This has been well documented in the last few years.

But a large part is less obvious: friction. We have the talent, the capital, and most of the technology. What we lack is a cheap, fast way to connect it all. I call that gap friction, and it’s what’s slowing our progress down.

Why friction? Progress on hard problems comes from iteration. You build something, put it in the world, watch what happens, and refine. Each turn of that loop teaches you something the last one couldn't. Run the loop fast enough and the learning compounds. Friction is what lengthens every cycle.

Consider what it takes to bring clean power online for a large company. The rules change with every region and every project type, the contracts pull in legal, finance, and operations. All of it has to clear before a single kilowatt-hour is generated, before you can start to learn, improve, and refine. It’s too complex, and that complexity causes friction.

How AI helps the climate

This is where AI comes in, and where most of the conversation about it goes wrong.

The complexity itself doesn't go away. The fragmented standards and opaque pricing are still there, but AI keeps them from being the bottleneck. For most companies, rigorously evaluating commitments across carbon, SAF, and renewable energy is out of reach today. They’re finding they can't solve it by adding more people, because that would increase the cost they’re looking to reduce. AI pushes that cost toward its floor, and that deflation is what scaling markets need.

Abstracted complexity doesn't mean fewer people. Hard judgment calls are demanded by a fast moving market, and those are the ones worth a person's time. Even though we’re at “day one” of the AI transition, we’ve already found a model that works across sectors and businesses: AI carries the complexity while experts make the decisions. That’s the job our embedded climate strategists do at Patch. Companies bring us in to extend what their own team can act on and to run the loop again and again.

None of this erases the fundamental cost: AI has a real and growing energy footprint. The largest US hyperscalers are on track to spend about $725 billion on data centers and AI infrastructure in 2026, up 77% from the year before, most of it on grids never designed for this demand. Sustainability leaders have to hold two things at once. They need AI to run their climate programs, and they have to account for what that AI costs the climate. The companies pulling ahead treat that as a single problem, one that’s better solved with thoughtful use of AI.

A future worth striving for

One thing I’m certain of: AI isn't incremental. It's improving faster than anything we’ve seen before and we're still early. That pace scares people, and it's also why it matters for the climate. We have very little time to make the progress the science calls for, and clearing the friction is how the iteration gets fast enough to make it.

We can have abundant clean energy, a breathable atmosphere, and a climate back in balance. We should also want every generation to expect more than the last. We can't keep trading one for the other. The only future worth striving for is a future of sustainable growth.

That's what we're building toward at Patch.