Concerning Data Centers

By Charlie Johnston

Our reader and commenter, Judy, brought up the growing controversy over data centers being sited all around the country. Working a political campaign, I realized just over a year ago this month, what a huge issue data centers were likely to become, so I have done a lot of study on the matter.

There are both short-term and long-term questions on data centers. Let’s look at the three major short-term questions first.

1. Power: Data Centers consume an enormous amount of power. On average, the electric bills of towns near data centers rise by 15-20% once the center comes online. The problem is compounded by the reality that in most places the increases keep coming, leading to spikes of as much as 267% over four years. While that is rare, spikes of 40% over the same time period are typical. If we are not in the infancy of data centers, we are certainly at the toddler stage. Governmental and Corporate promises to keep prices under control have, thus far, proven less reliable than their promises that the Covid shots were “safe and effective.” When you routinely get wildly inaccurate information from supposedly authoritative sources, you are a fool to put much faith in their claims going forward. Things could level out or spike wildly out of control. The “experts” are too often propagandists for you to know much of anything without doing substantial private research. For that, you need some specific skills. Otherwise, you risk jumping to indefensibly large conclusions from tiny amounts of data – and that only confuses things further. (Kind of like the supposed national “experts” routinely do when they are not outright lying.)
2. Water: Data centers are also very thirsty. They consume enormous amounts of water to cool their servers. In areas that have vulnerable aquifers, data centers can quickly come to compete with local residents for water. This can create both cost and availability problems. In the worst cases, it can cause aquifer depletion. No one really knows how much thirstier data centers might get (though everyone involved has an opinion – often masquerading as “facts”). On this subject, too, there have been wild claims that do not even get in the general vicinity of facts. I sat in a meeting of county officials where data company hopefuls were promising that their project would consume no more water per day than three new households in the community. That was utter, bald-faced, poppycock. And the county was taking the assurance at face value without demanding an enforceable performance mechanism.
3. Siting: In many areas, state legislatures irresponsibly made it open season for data center companies, allowing them to effectively bum-rush local communities that did not have effective means of regulating or stopping them. That could change the nature of a community overnight and create a major public nuisance rather than an opportunity for economic growth.

Short-term, the main problems with data centers have been and are resources and planning. There is plenty of land to site data centers without destroying the aesthetic benefits and character of local communities. In fact, there is plenty of land that sits adjacent to communities that, with good planning, can become economic engines rather than degrading eyesores that suck up precious resources.

In Texas, counties do not have zoning or regulatory authority over their unincorporated areas. That means that while towns, villages, and cities have some measure of control over how a data center can be sited and authority to regulate it, counties have no control except whether to offer a tax abatement. If a data center company wants to push forward regardless of any tax agreement, there is nothing the county can do about it.

Anywhere in the country, a municipality other than a county can annex land and then site centers right next to another municipality, protecting itself while damaging its neighbor. States must set up siting regulations. A municipality should not have sole authority to mount a project with profound regional consequences while denying those other municipal entities significantly affected any control over the process.

On the matter of power, President Trump has said that data centers must pay for their own electricity. I haven’t seen any statutory enforcement mechanism for that, though. Frankly, it is primarily a state issue, anyway. States must adopt clear rules.

Coastal states can set up desalination plants to provide sufficient water for data centers and to supplement existing water availability issues.  As of two years ago, a 25-million gallon per day desalination plant could be built and brought online for $60-90 million dollars. From approval to operation takes an average of three to five years. (If your local officials tell you it is longer, it is because of unnecessary hurdles those officials have mounted). In other areas of the country, pipelines can be built to ship abundant fresh water. The land the nation of Israel sits on was once forbidding desert, which Israelis made bloom through desalination plants and pipelines.

As for energy, those states (all blue) which refuse to build and site new power plants are out of luck. Data centers cannot be sited within them without depleting resources and making the blue states even more unlivable than they already are. Of course, that does not mean blue states will not site them. Ordinary people always come last in blue states, if they are considered at all. Even in states that will allow new power plants to go up, things are going to have to be dramatically increased if they are to keep up with the demand. Several universities around the country have ventured into work on thorium-based nuclear power plants. Thorium is more plentiful, but harder to mine and refine, than uranium. The best quality of thorium is that it can be refined to run a power plant, but cannot be refined to make a bomb. It is significantly safer than uranium-based power plants while offering all the energy benefits.

The short-term problems are fairly easily solvable if we have the resolve and will to do it. But voters MUST make their state and local officials accountable, or you will have a helter-skelter patchwork that will probably do more damage than good.

Passions are very high – and growing – right now. There is a lot of information that does not hold up when you examine it. Many of the companies proposing data centers are giving rosy calculations which bear little resemblance to reality. They are engaged more in marketing than informing.

I am well aware that any time any socially-transforming technology makes landing, there is a Luddite tendency to exaggerate the dangers so as to keep things just as they are. Reality is that things never stay just as they are. In the mid-1800’s many folks in the Midwest organized to stop train stations from being established in their communities because it would change their towns from what they liked about them.

In the west, ghost towns are primarily towns that grew up in gold rush areas and then died off when the gold rush ended. In the Midwest, ghost towns are those towns that successfully blocked train stations from their borders – and then died because they cut themselves off from what was the rapidly developing primary means of transport – both of people and of goods. Things never stay the same. People either adapt or die. If they adapt well and prudently, they thrive. If they do it helter-skelter, they may kill off their communities. But trying to preserve a community intact in amber is an almost sure-fire way to kill that community off. Root hog, or die.

While I would not take the Luddite approach of trying to freeze everything in amber just as it is, there are serious long-term questions about data centers that should be dealt with. It is very difficult to do justice to those because we are caught in the switches. Trump has declared that AI is a national security issue that we must lead in – or face terrible new threats from hostile nations. He is right – and data centers are the vital nervous system for AI. If we were to fall behind in AI, hostile nations such as China – and even very small nations – with a vigorous program promoting such centers could quickly pose an existential threat. If that happened, we would never even get to consider how to harness AI to its benefits while recognizing and discarding the chaff that comes with it – for within a generation, we would no longer be the autonomous “land of the free.”

Science fiction movies of the 50’s and 60’s (Them, The Beginning of the End, The Day the Earth Stood Still – and many others) were expressions of visceral fears about what the larger, long-term consequences of nuclear power would be. Reflexive opposition to data centers is an expression of exactly that sort of visceral fear of AI. That fear is perfectly reasonable – and prudent. Pope Leo XIV began to address those fears with his recent Encyclical, Magnifica Humanitas. Notice that I say he began the serious conversation on it, not that he resolved it. He laid out some critical principles, though, in what I think is a very good starting place. More’s the pity that few will pay any attention to it. But that is the price to be paid when the hierarchy emphasizes trendy political issues rather than focusing on faith and morals. The world does NOT start treating the hierarchy as politically astute because of its grab for political authority; it starts ignoring it even in its area of genuine expertise and authority.

I am not worried about any sort of Terminator-like takeover of the machines. I am very worried about how this will dumb people down. Forty years ago, if I knew your name I also knew your phone number by heart. Now I’m lucky to remember my own, because my phone serves as a crutch to remember them for me. When Google came on the scene I was appalled a few years later at the huge profusion of what I called, “Google intellectuals” – people who knew no subject deeply but thought they could wing it by looking it up on Google. Just read our social media sites and you will see “Google intellectuals” swarming like ants to a spilled honey truck. Having superficial access to a host of issues does not increase critical thinking. It just allows midwits to think they are geniuses without knowing much more than they did three decades ago – maybe even less. And they are cocksure of their midwit fantasies, because a Google search allowed them to find supporting evidence of their position while they ignored all contrary data.

We are becoming dumber, while thinking we are becoming smarter. The truth is we are become technological wizards while degenerating into philosophical and theological tribes of competing monkeys. AI will only advance that disorder.

I am also concerned that AI systems too often go rogue, making false statements and then creating support documents that do not and never existed – to prop up their poppycock. AI programs are subject to the same dictum earlier computer programs were: Garbage in, garbage out. That they can go rogue and falsify data and documents to support those false premises is a terrifying reality in a world that increasingly is unable to discern truth through rigorous critical thinking and research. Mankind’s future could be less Skynet and more Mad Max. It is ironic that unchecked AI could lead us into a new dark age instead of enlightenment.

The key is to understand that AI can perform tasks with lightning speed. It can make connections between vast troves of data. But it does not have the spark of innovation, creative ingenuity, and deep insight beyond what data sets show. That is a human capacity. My main fear is that AI, in performing tasks brilliantly, simultaneously diminishes the creative capacity among the only species that has it in an advanced way, by falsely convincing them they don’t need it.

The short-term questions are resolvable, though I doubt we will do so in any comprehensive, coherent way. That is not terrible. Most major advances have come in fits and starts, only refined over time. The long-term questions have not yet been seriously tapped. They are complicated by the actual need for AI in military applications to insure our short-term survival. If we don’t survive as a free people, we won’t be able to worry long-term about AI at all.

Whenever I see such a conundrum, I ask myself whether God is intentionally herding us for His purposes. Right now, it feels an awful lot like we are being pushed into a cattle chute. If I did not know that God intends our reclamation, not our destruction, I would be very worried, indeed.

*********

The air conditioner on my car went out last week. Apparently, it happens to a lot of folks this time of the year in Texas, so I had a two week wait before I could get an appointment to get it fixed. Initially, I thought, “What the heck?” I was almost 30 years old before I had AC in my vehicles anyway, so it would be a perhaps nostalgic return to an aspect of my long lost youth. I quickly learned two things:

1. Driving in the Chicago area without AC in the summer is an entirely different thing than driving around Texas without AC in the summer.
2. Driving without air in your late 20’s is an entirely different thing than doing it when you are 70.

Once repairs are finished, I will be more grateful for AC than ever. (It would not surprise me if a lot of European World Soccer Cup tourists go back home demanding AC in their cars, too. Go, Freddy!)                       

If communication goes out for any length of time, meet outside your local Church at 9 a.m. on Saturday mornings. Tell friends at Church now in case you can’t then. CORAC teams will be out looking for people to gather in and work with.

Find me on X at @JohnstonPilgrim

Donate to CORAC!

Join the Conversation!

The Corps of Renewal and Charity (CORAC)

18208 Preston Rd., Ste. D9-552

Dallas, Texas 75252