Before We Say Yes or No to a Data Center, What Should We Ask?

I have been trying to understand the economics of data centers. So far I have learned that they cost a great deal to build, that they are expected to make a great deal of money, and that the people discussing the first number are not necessarily talking about the same thing as the people discussing the second.

I would like to know what happens in between.

I think AI could create considerable economic value. There are things businesses can do with it now and probably things they haven’t thought of yet. I also think demand here and competition abroad will keep companies looking for places to build the computing capacity they need. But neither conviction tells me what to think about a particular project in Northern Nevada.

For that, I’d like to ask a few questions. Some are about what we might have to provide. Just as many are about what we might get.

Who Is Going to Pay to Use All This Computing Power?

When I read that trillions may be spent on AI infrastructure, I wonder where the revenue will come from. I pay for some technology subscriptions now, but I don’t think my household is going to carry its share of the next trillion. There are limits, even if my inbox doesn’t appear to recognize them.

Some of the money will come from people paying directly for AI services or buying products with AI built in. Businesses might pay for it out of budgets they already have. If a distributor can use AI to reduce shipping errors and the resulting refunds and rework, it can pay for the tool out of the savings. Nobody has to charge the customer more for that to work. Of course, the distributor needs to count what it spends integrating the tool and checking its work. Otherwise, it may just be moving the expense to a department that wasn’t invited to the meeting.

The more interesting possibility is new business. What can someone now make, sell, or deliver that used to be too expensive or impractical? If enough companies find good answers, we could see more economic growth over time. Building the facilities is one kind of economic activity. Finding customers for things made possible by those facilities is another, and it’s the one that has to keep going after the construction crews leave.

What Might Stay Here?

Nevada’s economic development board considered a Vantage data center proposal in Storey County that projected about $254 million in equipment investment and 10 direct full-time jobs within the first five years of operation. The projected average wage was $64.20 an hour. Those are good wages for the people who earn them, but there aren’t many chairs in that particular break room.

That doesn’t make it a bad deal. The state’s analysis projected substantially more economic activity from construction than from operations. That makes me want to know how much of the construction work goes to Nevada tradespeople and contractors, and what local companies might keep selling to the facility once it opens. I’d also like to know what the county expects to collect in taxes after incentives, and when. A large investment number is a good reason to ask those questions. It doesn’t answer them.

I’m just as interested in what we could build around the technology. At the University of Nevada, Reno, researchers are developing a robotic watering system for sheep. One idea is that moving the water could change where the sheep graze and reduce wear on the pasture around fixed troughs. I don’t know whether the sheep have been consulted, but I can see why a producer would be interested. The next questions are whether it works well enough on the range and what a rancher would pay for it.

That project is research, and I have no evidence it depends on a data center in Storey County. I bring it up because it shows where some of the larger economic promise might come from: people here using new tools to solve problems they understand. If we want some of that value to stay in Nevada, what would help? Access to computing for local firms? Training tied to actual work? Contracts that give local suppliers a chance to earn repeat business? Those are opportunities I’d like to hear more about.

Who Gets the Utility Bill?

I’d ask where the electricity and water will come from, how much the facility expects to use, and who pays for any new infrastructure. If a developer says it recycles water, I’d like to know how much fresh water it still needs. If the utility builds new capacity for a project that later shrinks or never opens, I’d like to know who pays for what was built.

These questions are sometimes treated as objections. I think of them as questions I’d ask before agreeing to almost any large purchase. If somebody tells me the price doesn’t include installation, I generally ask what installation costs.

Could We Check Back in a Few Years?

I’d be more inclined to trust the projections if I knew which ones we could check later. How many people were hired? How much work went to local companies? How much tax revenue came in after the incentives? Did the facility use the water and power it expected to use? If the results differed from the projections, what did we learn?

I started out wondering who would pay for all this computing power. I still want to know. But there’s a question closer to home that we have a better chance of answering: what would make hosting some of that capacity worthwhile for Northern Nevada? I can imagine a good deal. I’d like to see one explained clearly enough that we could recognize it.

If a data center were proposed near you, what’s the first question you’d want answered before making up your mind?