An industrial-symbiosis model, borrowed from Denmark, for turning AI’s power demand into shared local infrastructure
Montana has always been a state that thinks in terms of rivers, forests, railroads, dams, and wide-open spaces. We understand infrastructure because we’ve spent generations building and living alongside it.
Now artificial intelligence is creating demand for a new kind of infrastructure: data centers. Around the country, companies are searching for locations with abundant electricity, water, land, and fiber-optic connectivity.
Instead of asking whether Montana should compete for these projects, perhaps we should ask a different question.
What if we reimagined what a data center could be?
Rather than viewing it as a giant warehouse full of computers, imagine it as the centerpiece of an industrial ecosystem.
Hydroelectric power from facilities like Hungry Horse Dam or Seliš Ksanka Ql’ispé Dam could support advanced computing. The heat generated by those computers, instead of being discarded, could warm commercial greenhouses, research facilities, aquaculture, or nearby businesses. Water could be recycled through multiple uses. Shared infrastructure could support agriculture, technology, manufacturing, and education together.
This isn’t science fiction. The Danish city of Kalundborg has spent decades building what is known as an industrial symbiosis, where one company’s waste becomes another company’s resource. Its success wasn’t based on a single revolutionary project. It evolved through practical partnerships that saved money, reduced waste, and strengthened the local economy.
Montana has an opportunity to learn from that model while creating one of its own.
Imagine an innovation campus where renewable energy, artificial intelligence, controlled-environment agriculture, workforce training, university research, and local businesses are intentionally connected rather than isolated. Such a project would require serious study, environmental review, tribal consultation where appropriate, utility participation, and community input. It may or may not prove feasible at any specific location.
But the conversation itself is worth having.
Artificial intelligence will require enormous amounts of electricity wherever it grows. The question is whether Montana simply watches that investment happen elsewhere, or whether we explore ways to capture more of the economic value while putting wasted energy to productive use.
The future may not belong to communities that simply produce power.
It may belong to communities that learn to use every watt twice.
I’d love to hear from engineers, utility workers, growers, tribal leaders, educators, and anyone else who thinks this idea deserves further exploration.
Glenn Phillips
Whitefish