I use generative artificial intelligence programs like ChatGPT and Google Gemini as tools sparingly. I did find some good uses for the software despite this, whether it is deciphering dialogue through recordings, looking for essay ideas or getting research results quickly.
For me, inaccurate information and art theft aren’t the only things scaring me away from the software. It’s the fact that every use comes at the cost of destroying the environment due to data centers being built around the United States, keeping the software up and running.
I need to set the stage on the impact these centers are already making on the environment, and the potential of the climate getting worse. For example, the overuse of water could make life a lot harder for people to survive if large data centers keep being built.
Miguel Yanez-Barnuevo, project manager for the Environmental and Energy Study Institute, wrote an article detailing the water consumption of data centers in June of 2025. He puts into perspective how much water is used by medium- and large-sized data centers.
To make sure the computer hardware running the AI software doesn’t overheat, these centers run water to keep the machinery cool. As described by Barnuevo, a mid-sized data center can use up to 110 million gallons of water per year, while a large-sized center can use five million gallons in just one day.
“Each 100-word AI prompt is estimated to use roughly one bottle of water (or 519 milliliters),” Barnuevo said, referencing a study from the University of California, Riverside.
That might not mean much, but for reference, five million gallons of water can provide running, clean water for a town of 10,000 to 50,000 people for one to eight days!
Barnuevo warned readers that water consumption by AI data centers is affecting communities.
He stated, “All data centers in Northern Virginia consumed close to 2 billion gallons of water in 2023, a 63 percent increase from 2019.”
He added that Loudoun County accounted for 900 million gallons in the same year, in turn making the data centers rely on drinking water for cooling the machines rather than reclaimed water.
Recent events have put data centers and their resource consumption into perspective.
For some odd reason, billionaire Kevin O’Leary recently decided to back a “hyperscale” data center in Utah, dubbed the “Stratos Project,” that would likely double the use of electricity and water in the state.
In a separate article for The Salt Lake Tribune, Leia Larsen talked with Utah State physics professor Robert Davies about the potential heat resulting from the new data center.
Davies predicts the nine gigawatts of energy that will be used to power the facility will produce another seven to eight gigawatts of “waste heat” produced by the data center to be released into the environment. The heat released into Hansel Valley will be the “equivalent of about 23 atom bombs worth of energy dumped into this local environment every single day.”
Larsen adds that Davies “predicts dumping that much heat and energy into Hansel Valley will raise local temperatures five degrees Fahrenheit during the day and up to 28 degrees at night.”
It frightens me how high the risks are, yet billionaires who insist on AI being the future do not care about how the people living in proximity to their data centers will be affected, even though they know the climate will get worse and water will eventually run out as a result of them.
While the technology has been used for good, the damage cancels everything out, especially since the damage, for just one data center, can dump 23 atom bombs onto a valley.
Data centers should be torn down if they have to use this much energy producing slop, especially if humanity is at constant risk of boiling alive.
