AI for Montana

Four data center proposals are live in this state. Our utility has preliminary agreements that could add 1,400 megawatts of new load by 2030, against a current annual load of roughly 750. That is not growth. That is doubling Montana's electricity demand for a single industry, in a state where only 16 percent of us are using these tools at all.

The position

Montana supplies the land, the power, the water, and the minerals. Someone else captures the value.

Sheep Creek is seeking fast-track federal permitting for critical minerals. There is an active proposal to pull rare earths out of the Berkeley Pit. The physical supply chain behind AI runs straight through this state, and we are not equipped to use what gets built here.

We have done this before with copper.

A state can host an industry and still end up poorer for it. I am not proposing that Montana refuse this one. I am proposing that we be positioned when the decisions get made. Fluent enough to use these tools. Organized enough to negotiate. Clear enough about what we want that we recognize a good deal when we see one.

Positioning, not predicting

What is actually at stake

Strip out the megawatts. Here is what this looks like on the ground.

Your power bill

Carnegie Mellon researchers found data center growth could raise electricity bills 8 percent nationally, and as much as 25 percent in some regional markets. Montana ratepayers absorb the cost of infrastructure built for someone else's servers.

Your water

A single large data center can use up to five million gallons a day. In a state where water rights are already the sharpest fight in the room, that is not a small ask.

Your county

A proposal in Bonner collapsed this summer after sustained public opposition. Missoula County is rewriting its rules. Elsewhere, agreements get signed under nondisclosure and residents find out afterward.

Your business

If you run a company here and your competitors two states over are using these tools well, the gap compounds every quarter. That is happening now, quietly, with no announcement.

None of that requires a prediction about the future. All of it is underway.

Four pillars

The same four words mean different things depending on where you stand.

Fluency

Practical skill. Being able to use it on the work you already do.

Agency

An informed voice. Knowing the tradeoffs well enough to argue about them.

Sovereignty

Who decides. Agency is being heard. Sovereignty is holding the pen.

Safety

Protection that holds no matter who is deciding.

Fluency comes first. People cannot argue meaningfully about a technology they have never had in their hands. Agency follows from fluency. Sovereignty is the door that fluency and agency make visible. Safety is the floor under all of it.

Three fronts

The pillars run across three fronts. Each one is active right now.

Practical AI

The fastest way to make Montanans harder to steamroll is to put these tools in their hands.

Hands-on sessions built around problems people already have. Led by local people, not outside experts. Platform-neutral, so the skill outlasts any one vendor.

Policy & governance

Montana has the pieces of an AI policy conversation. It does not have the conversation.

406 JOBS is real. The university AI institutes are real. None of them puts entrepreneurs, county officials, utilities, tribal governments, and legislators in the same room. That room is what I am proposing to build.

Physical AI

My position is not for or against development. It is that Montana deserves a benchmark.

Communities have no way to judge whether the deal in front of them is good or bad. No shared standard exists for what proportional community benefit looks like. That gap is fillable without taking a side on any single project.

Clock running

What this needs

I have a framework, a region, and a method. That is not an organization yet.

In Northwest Montana I already run a monthly AI meetup and a lab series for entrepreneurs and small business owners. That region is the proving ground, not the point. Here is where I need people.

Energy and water policy

The physical AI work is the most urgent and the place I am least expert. I need people who understand utility regulation, water rights, and permitting in this state.

Operators inside companies

Helping businesses integrate AI safely is the thinnest part of this framework. It should not be. I want people who have actually done it.

Policymakers and staff

Not endorsements. People who will tell me honestly what is realistic before the 2027 session and what is wasted effort.

A second region

The test of whether this scales is somebody outside Northwest Montana running it, with local people and a local face.

Tribal engagement, done properly

Montana has eight tribal nations and seven reservations. Sovereignty is not a metaphor here. I want a real data sovereignty process before doing work that touches it, and I am not the right person to design that alone.

People who disagree with me

Especially on data centers. A framework that only convenes people who already agree is not a framework. It is a mailing list.

Get involved

Montana does not get to decide whether AI reshapes this state. We get to decide whether we had a say in how.

If you read this and recognized something you have been trying to name, tell me. If you see a piece of it differently, that conversation needs more than one side in the room.

Takes a minute. No list rental, no newsletter blast. This does not work as one person's project. It works if enough of us decide Montana should have a say in how this goes.