How the AI boom could finally begin paying its way locally - starting in Bradford: Mark Lee

The problem with most data centres is not that they exist. It is that they waste something valuable. Heat.

Every time we ask ChatGPT a question, electricity is turned into computation and heat. The computation is prized. The heat is treated as a nuisance.

That would be merely wasteful if we were awash with warmth. But we are not.

Across Yorkshire homes, pools, leisure centres and public buildings are struggling with high heating costs and volatile energy prices. We have places short of heat sitting next to infrastructure throwing vast amounts of it away. That is not a quirk of physics. It is a failure of design.

There is no technical reason why the heat generated by data centres cannot be captured and reused locally. When you do that, digital infrastructure stops being purely extractive and starts behaving like civic infrastructure. The electrons are used twice. The same unit of energy does more work.

That is the thinking behind what we are planning in Bradford.

We are developing a high density data centre designed from the outset to connect into a local heat network. Instead of venting heat into the sky, it will supply it to nearby buildings. Bradford gets cheaper bills, we get efficient cooling. Everyone wins.

We are confident that our facility in Bradford could significantly cut heating costs each year. That is money that would otherwise be spent on gas, exposed to volatile prices, flowing out of the local economy. Done at scale, this has three outcomes that matter to communities.

First, it reduces energy costs. Heat recovered from computing is effectively free at the point of use. Over time, that lowers bills for the buildings connected to it and

reduces exposure to volatile gas prices.

Second, it cuts carbon. Reusing heat means burning less fuel elsewhere. It is not a silver bullet, but it is immediate, measurable and permanent.

Third, it changes the relationship between digital growth and place. Infrastructure that visibly benefits its surroundings earns its licence to operate. It becomes something communities want nearby, not something imposed on them.

Bradford is not unique. What works there can work across Yorkshire and well beyond it.

This matters because the demand for compute is not slowing down. AI will continue to expand into healthcare, logistics, science and public services. Trying to stop that growth is neither realistic nor desirable. The real question is whether we build the infrastructure of that future in a way that deepens existing problems or helps solve them.

We need to stop treating digital systems as abstract global assets and start designing them as part of local economies. If we get it right, the AI boom does not have to be something that happens to Yorkshire. It can be something Yorkshire leads and benefits from.

Mark Lee is CEO of Deep Green