We Can Live with AI, but Not Like This

Melissa A Tamberella

Jun 12, 2026

Nobody asked us. That is the part that keeps getting lost in every conversation about artificial intelligence and what it is doing to our jobs, our water, our electric bills and our communities. Nobody held a town hall. Nobody put it on a ballot. Nobody sat down with the people whose lives were about to be turned upside down and said here is what is coming, here is what we are going to do about it, and here is how we are going to make sure you are not left behind. The technology just arrived, the data centers just got built, the jobs just started disappearing, and everyone in power kept saying trust us, the benefits are coming. We are still waiting on the benefits. The costs showed up right on schedule.

I have been writing about this all week and every piece connects back to the same question. Can humans and artificial intelligence actually coexist in a way that works for everybody and not just the people who already have everything? The answer is yes. But not the way we are doing it right now. Right now we are doing it the worst possible way - full speed, no guardrails, no plan for the people getting displaced, no transparency for the communities getting a data center dropped in their backyard, and no honest conversation from anybody in power about what the finish line actually looks like. That has to stop and here is what has to replace it.

WAIT - WHY ARE THEY SCATTERED EVERYWHERE IN THE FIRST PLACE? ────────────────────────────

A lot of people are asking the same question I had - why are these things being built all over the place instead of in one location away from everybody? It is a smart question and the answer will make you angry because none of the reasons have anything to do with what is best for the communities absorbing them. There are four reasons and every single one of them is about money and speed.

First is power. Every data center needs an enormous amount of electricity hooked up immediately and no single location in America has enough available grid capacity to handle thousands of them at once. So companies fan out across the country hunting for wherever the power is available and the local government is easiest to work with. Second is speed - the AI race between Microsoft, Google, Amazon, Meta and the rest of them is moving so fast that nobody is stopping to ask whether a location makes sense for the people living there. They find available land, available power and a local government they can work with quietly and they break ground before anyone knows what happened. Third is tax breaks - states and counties are in a bidding war offering billions in incentives to attract these facilities, so companies shop around for whoever gives them the sweetest deal and moves fastest. Virginia got so overrun it is now the most concentrated data center region in the entire country and its own residents are fighting back hard. Fourth and most important - nobody is in charge. There is no federal agency coordinating where these go. No national plan. No zoning authority at the federal level telling companies they cannot build next to a neighborhood school or pull water from a drought-stressed county. Every company just picks wherever works for them and moves before anyone can stop them.

Sixty seven percent of all new planned data centers are going into rural areas. Thirty nine percent are going into counties that currently have zero data centers - meaning communities with no experience, no existing regulations and no idea what is about to hit them. The South alone accounts for nearly half of all planned construction. Working class and rural communities are being targeted because the land costs less and the assumption is that the people there have less power to fight back. That assumption is wrong and the communities proving it wrong every week in packed council chambers across the country are the reason $98 billion in projects got blocked or delayed between March and June of last year alone.

THE SMARTER WAY TO DO THIS

────────────────────────────

Here is what a rational national approach would look like. Designate federal data center zones in areas with abundant renewable energy, access to water that can actually support the demand, distance from residential neighborhoods, and existing infrastructure that can handle the load. The desert Southwest has solar power and space. The upper Midwest has wind. Parts of the Pacific Northwest have hydroelectric. Build the massive training facilities - the ones that consume the most power and water - in places designed and resourced to handle them, with federal oversight, federal environmental standards and real community benefit agreements that put money back into the host region. Keep the smaller inference centers - the ones that just deliver AI responses to users - closer to population centers where they need to be for speed, but with strict local zoning, water limits, noise requirements and mandatory public approval processes before a single permit gets signed.

Aurora Illinois showed exactly how this works at the local level. They put a 180 day pause on all new data center approvals and used that time to write actual standards - noise requirements, water consumption limits, energy requirements, and a rule that every new project has to go through full city council approval so nothing gets signed without the public knowing about it. Wisconsin lawmakers introduced legislation banning NDAs between data center developers and municipal officials after reporters found five towns had been legally gagged before their own residents knew what was being proposed. More than 30 states have introduced over 300 bills in 2026 dealing with data center regulation. The momentum exists. A federal framework to back it up and set a floor that every state has to meet is the missing piece.

FIX THE RESOURCE PROBLEM WITH TECHNOLOGY THAT ALREADY EXISTS ────────────────────────────

The water and energy problems are real but they are not unsolvable. Researchers have found that the massive amounts of waste heat data centers produce just by running can be captured and used for district heating, water purification and direct air capture of carbon dioxide. Instead of dumping that heat into the atmosphere and draining municipal water supplies to cool it down, facilities can recover 70 to 80 percent of their water using reverse osmosis recycling systems that already exist and already work. A data center in Fayette County Georgia was secretly pulling 29 million gallons of water through connections the county did not even know existed - the residents only found out when their water pressure dropped at home. That is not a technology failure. That is a transparency failure and a regulatory failure, and both of those are fixable right now without waiting for anyone to invent anything new.

NOW THE HARDER ONE - THE JOBS ────────────────────────────

Anybody who tells you AI is not displacing jobs is not being straight with you. An MIT and Boston University report found AI will replace as many as two million manufacturing workers by 2026. McKinsey projects that by 2030 at least 14 percent of workers globally will need to change careers entirely because of AI and automation. The white collar hits are already landing - customer service, paralegal work, bookkeeping, marketing coordination, technical writing, administrative support. These are not jobs at the bottom of the wage scale. These are the jobs that built the American middle class and they are being automated faster than any retraining program is currently equipped to handle.

But here is what the doom conversation leaves out. AI is also creating jobs and will create more of them. Data science, cybersecurity, AI ethics, machine learning, prompt engineering, AI system maintenance, human-AI collaboration roles that did not exist five years ago - these fields are growing fast and not all of them require a four year computer science degree. What they do require is access to training, which is where the system is currently failing the people who need it most. Reskilling programs exist but they are inconsistent, underfunded and inaccessible to workers in smaller cities and rural areas who are getting hit the hardest. That is a solvable problem if anyone in power decides to treat it like one.

WHAT IS ALREADY WORKING SOMEWHERE ELSE ────────────────────────────

Denmark and Switzerland have been handling industrial displacement for decades through wage insurance programs that catch workers when their jobs disappear and carry them at close to their previous salary while they retrain. The worker does not fall off a cliff. The family does not lose the house. The community does not hollow out. Germany ran a similar model and its manufacturing workforce absorbed enormous automation over the past 20 years with far less devastation than American workers experienced during the same period. These are not ideas from a policy paper. They are running systems in countries that decided workers were worth protecting and built the programs to prove it.

On the tax side several economists have proposed a token tax - a small levy on every AI model call, meaning every time a company uses AI to do something a human used to get paid to do, a fraction of a cent goes into a fund that supports the humans being displaced. The collection point already exists at the cloud provider level. Microsoft, Google and Amazon are already billing companies for AI usage by the token. Adding a small displacement tax on top of that billing is not complicated. Norway built a $1.5 trillion sovereign wealth fund on North Sea oil revenue and used it to give every citizen a stake in the country’s resource wealth. The United States is sitting at the center of the most profitable AI buildout in history and has not collected a single dollar of it for the public absorbing all the disruption. That is not inevitable. That is a choice being made by the people writing the rules, and choices can be changed.

WHAT YOU CAN DO RIGHT NOW ────────────────────────────

Go to brockovichdatacenter.com and look at the map. If a data center is planned near you, find out what your local zoning says and whether your officials have signed any NDAs. Show up to council meetings. Run for your local zoning board. Support candidates at every level who are willing to say out loud that technology companies do not get to rewrite the rules of local democracy because they have more lawyers than your town does. Call your representatives - the Florida numbers are in my previous post - and tell them you want data center transparency legislation, water consumption standards, a ban on secret deals between developers and elected officials, and a federal framework for where these facilities get built.

Learn enough about AI to use it instead of fear it. I use it every single day and I correct it constantly because it makes mistakes and cannot think for itself the way a human can. It is a tool - a powerful one but still a tool. The people who learn to work with it rather than against it are going to be better positioned than the ones who pretend it does not exist. The goal is not to stop the technology. The goal is to make sure the people who built this country and keep it running every day get a seat at the table where the benefits are being divided up, because right now that table has a very short guest list and none of us are on it.

────────────────────────────

The AI future is not something that is happening to us. It is something we get to shape if we decide to show up. The only question is whether we show up before the decisions are already made or after - and in my experience after is always too late.