When Big Tech Meets Small-Town Politics: The Data Center Dilemma TLWH 10.16.26

As a younger business owner, I found that the latest technology helped us build a growing, sustainable operation. Fax machines helped us eliminate error-prone issues that arose when we transmitted information over the phone. One of our customers, a Fortune 500 company in the Twin Cities, adopted voicemail, which made scheduling periodic sales calls more efficient. Email and Electronic Data Interchange sped up business processes. I adopted the BlackBerry phone, which let me carry email and voicemail in my jacket pocket. 

If I thought about Artificial Intelligence (AI) then, I believed I would be moldering in my grave before I’d have to think about it. 

But when the writing demons took over, I found that spell checkers that caught spelling, punctuation, and grammar errors were probably AI. Now, even cars use AI to track our location, keep us in our lane, automatically dim our high beams, and even parallel park. 

Mathematician and computer scientist John McCarthy coined the term "Artificial Intelligence" in 1955. It ain’t new. High-profile successes like ELIZA (the first conversational chatbot built in 1966) led researchers to predict that human-level AI was only a few decades away.

In 1997, IBM’s Deep Blue defeated world chess champion Garry Kasparov, proving AI could beat human tactical reasoning. In 2011, IBM’s Watson defeated human champions on the quiz show Jeopardy!, demonstrating advances in processing natural language queries.

It’s hard to imagine the stress Google’s efforts to build four AI data centers on 280 acres of rural land on the west side of the city put on Hermantown's mayor and city councilors. As important as those city officials are, I bet they never expected to meet the juggernaut that is Google, along with its contractors and developers. The arm-twisting from them on one side and the grassroots pushback from the other must be causing real heartburn for the electeds.

Where we live, you don’t have to go far to see the Stop the Hermantown Data Center lawn signs. Local news sources and social media pundits are keeping the issue front and center. As the development process proceeds, the elected council is pressured on one side by wealthy developers and trade unions, while the NIMBY efforts of real people who live here push back. Many people will be seriously unhappy with the elected officials' decisions. They can’t win.

Recently, a 27-year-old British AI researcher, Jacob Coxon, appeared on Jon Stewart’s The Daily Show. He is a former AI researcher who worked at OpenAI and Anthropic, two of the biggest players in artificial intelligence, and is considered a whistleblower after he made his resignation letter public. In his letter, he warned, “They are racing to develop self-improving superintelligence while gambling with our lives."

Stewart asked Coxon what AI developers see as AI's ultimate goal. According to the industry insider, these creatively financed, opaque corporations see AI's ultimate goal as creating a never-before-realized utopia.  They claim AI will, after eliminating most jobs and replacing workers with technology, lead to an "age of abundance" where no one needs to work and everyone receives a universal basic income. Everyone has world-class health care—a retiree’s dream.

It also sounds a lot like the communist dream, except in this AI scenario, a relative handful of people will own all the capital, not the people’s government. 

Something I didn’t know: our country has roughly 5,000 active data centers operating today, about 40% of the world’s data center footprint. Most standard facilities are smaller enterprise or regional colocation centers. Hyperscalers like Amazon, Microsoft, and Google drive the cloud backbone.

AI workloads require higher power density and liquid cooling, so tech companies are building dedicated AI mega-campuses like the proposed Hermantown one rather than traditional cloud facilities. Over 780 data centers are currently under active construction, and roughly 3,700 additional projects are announced or in the planning stages across the U.S. 

Industry estimates say nearly half of planned AI data center builds face delays or cancellations because of power grid limits; local political opposition over noise, water, and electricity use; and multi-year supply chain backlogs for hardware like transformers and turbines.

My BlackBerry is a classic antique. It only took a few years for the iPhone and Google Android phones to make it obsolete.

The chips inside today's data centers have a surprisingly short primary lifespan of two to three years. The transformers, liquid-cooling loops, and backup power generators built into today's facilities have a physical lifespan of 10 to 15 years. Data centers don't get abandoned when chips age, though. Like we humans, operators systematically swap out internal racks and retrofit systems every few years.

So, will AI take over society and create a Shangri-La of bliss and harmony? More importantly for me, will I be around to find out?