AI Data Centers Reimagined

Data centers are being sited faster than communities can form a view about them. The public conversation has settled into two positions — build them because the economy demands it, or block them because they take water, power, land, and give back a handful of jobs. Both positions accept the same premise: that the building is what it is.

We think that premise is the weak point. A data center is a large, capital-rich, thermally active, permanently staffed structure placed in a specific watershed. Almost nothing about its current form is required by physics. It is the product of a procurement culture optimizing for speed and cost per megawatt, and nothing else.

So the question we are working is narrower and more useful than a verdict: what would the building have to become before a community would genuinely want it there?

The next decade will place an enormous amount of new computing capacity somewhere. Where it lands, it will set land prices, water agreements, grid investment, and local politics for a generation. These are effectively permanent decisions being made under time pressure with very little public design work behind them.

Rural western communities are especially exposed. They have the land, the transmission access, and the tax appetite, and they usually lack the technical staff to negotiate as equals. The result is a boilerplate agreement written by whoever arrives with a template.

A single well-documented alternative changes that dynamic more than any amount of opposition. If a county can hold up drawings, costs, water figures, and a measured benefit standard from a comparable project, the negotiation starts from a different place.

We are gathering accounts from people who live inside this question already: residents near existing facilities, county commissioners and planners who have voted on them, utility staff, water managers, ranchers whose land borders proposed sites, and technicians who work inside the halls.

What we are hearing so far, and want to hear much more of: noise and light are described more often than water; the jobs question is more contested than the energy question; and almost every community says the same thing about process — they learned about it late.

We are assembling the technical and regulatory record: cooling and closed-loop water literature, waste-heat recovery precedents in Northern Europe, hydrothermal carbonization research, agrivoltaic and solar-canopy structural studies, microgrid interconnection rules, Colorado water law, and county-level siting agreements already in force.

Every source is listed with its provenance and its limitations. Where studies disagree — particularly on heat recovery economics and on real versus nameplate water use — we record the disagreement rather than choosing the convenient number.

These are the concepts currently growing side by side. None is settled. Several may not survive the bench.

This is where the project sits today. The bench work underway covers structural cost of the canopy against its energy and durability return; realistic closed-loop makeup water at Colorado altitude and humidity; the honest economics of low-grade heat when the receiver is seasonal; interconnection and tariff treatment for a microgrid that exports; and the permitting path through county land use and state water administration.

Two assumptions are currently the most fragile, and we are saying so plainly: whether heat recovery pays for itself without a subsidy, and whether a canopy structure survives cost review against conventional roof-mounted arrays. If either fails, the design changes rather than the claims.

We are forming the founding team and seeking partners: a mechanical and water engineering firm, a utility or rural electric cooperative willing to co-design interconnection, a landowner or county open to hosting a study site, a grower interested in operating greenhouse capacity, and funding for a rigorous feasibility study rather than a rendering.

Governance is being drafted under cooperative principles through Harmony Development Cooperative, so that the community hosting a pilot holds a real and durable stake in it.

The intent is a modest first facility in Colorado — small enough to build honestly, instrumented thoroughly, and open to visitors. Not a showcase. A working building whose numbers are published whether or not they flatter the design.

If it works, the useful output is not the building. It is the drawing set, the cost record, the measured performance, and the benefit agreement — released openly so the next county has something to negotiate with.