How AI is harming Wisconsin’s path to renewable energy
Despite public opposition, Wisconsin's Public Service Commission approved We Energies' proposal to build two new fossil fuel-burning power plants, in part to meet demand from data centers and AI.
The Recombobulation Area is a thirteen-time NINETEEN-TIME Milwaukee Press Club award-winning opinion column and online publication founded by longtime Milwaukee journalist Dan Shafer. The Recombobulation Area is now part of Civic Media.
We’ve long been primed by popular culture to expect some kind of robot apocalypse. From the cyborgs of the “Terminator” franchise to the rebellious artificial intelligence, Hal 9000, from “2001: A Space Odyssey,” the authors of science fiction have imagined a wide variety of ways in which our own technologies may turn on us and ruin the planet.
I don’t think I’ve read or watched anything that managed to predict our current situation, though. The rapid rise of AI and its associated hype is creating consequences for everything from how we learn, how we work, and — recently relevant to Wisconsinites — how much energy we consume and where it comes from. It may be a far less cinematic form of destruction, and one that people largely seem to embrace or at least be indifferent to, but it’s just as insidious.
Despite overwhelming opposition from businesses, local governments, environmental groups, faith leaders and community members, the Public Service Commission (PSC) recently approved We Energies’ proposal to build two new fossil fuel-burning power plants in Wisconsin (in Milwaukee and Kenosha counties). They’ve cited the construction of massive new data centers by tech companies like Microsoft and the increased energy demands from them as a major reason for the move. One of the major reasons cited for the expansion of data centers is the proliferation of generative artificial intelligence tools, like ChatGPT and Co-Pilot.
Every single time we use AI tools to create a cover letter, a code, or a script for breaking up with our partners, energy is consumed. The computing power required to run generative AI is, in fact, massive — like the profits they’re generating for tech companies.
The PSC’s approval, for instance, comes despite the sure knowledge that building expensive new gas power plants locks us into using non-renewable, polluting energy sources at a moment when time is of the essence if we have any hope of slowing the impacts of the climate crisis. While the new gas plants will create less pollution for the surrounding communities than old coal-fired plants, the risk factor to people’s health is hardly zero.
According to Clean Wisconsin, “pollution from the plants could cause between $80 million and $127 million in health costs for the state each year from respiratory and cardiovascular disease.”
All of this further shows that, without meaningful and binding regulations on both energy companies and the corporations driving demand, it’s all too easy for them to walk away from previous “commitments” to doing the right thing. Tech giants like Google, Microsoft, and Apple have admitted in recent sustainability reports that AI is largely responsible for driving up their energy use. Google’s greenhouse gas emissions, for instance, have increased 48 percent since 2019, all but negating the company’s goals of reaching net zero by 2030.
Read more from Emily Mills at The Recombobulation Area here, and subscribe to their newsletter on Ghost.
Meanwhile, most of the hype around the benefits and capabilities of generative AI is wildly overblown, when it’s not an outright lie. Its biggest proponents are profiting off the stolen works of artists, writers, coders, academics, and others who never approved the use of their materials in training AI.
No matter how you feel about it, generative AI has become nearly unavoidable, too — from the automatically generated AI summaries at the top of your Google searches, to ChatGPT, to AI “assistants” installed on our phones. And every single one of those uses sucks up energy, more and more every year.
According to a recent article in Wired, which has covered AI and its energy costs extensively:
“Last month, the International Energy Agency released a report finding that data centers made up 1.5 percent of global energy use in 2024—around 415 terrawatt-hours, a little less than the yearly energy demand of Saudi Arabia. This number is only set to get bigger: Data centers’ electricity consumption has grown four times faster than overall consumption in recent years, while the amount of investment in data centers has nearly doubled since 2022, driven largely by massive expansions to account for new AI capacity. Overall, the IEA predicted that data center electricity consumption will grow to more than 900 TWh by the end of the decade.”
Setting aside the many real questions and concerns around the value and ethics of generative AI, the energy costs alone should give everyone pause, especially as they result in negative impacts on our communities. Regular people are paying the price, especially as so many utility companies rush to build more fossil fuel-burning plants to meet the associated demand. They should instead be making meaningful investments in the large-scale renewable sourcing, like solar and wind, that we need to avert the worst health and environmental outcomes.
The solutions are available to us, if we have the will to throw a wrench in the machine, slow down, and demand it. For our sake, and for the sake of every future generation.
Emily Mills is a longtime freelance writer/reporter based in Madison. They previously served as Editor of Our Lives, Wisconsin's only LGBTQ+ media outlet, and as an opinion columnist in the Milwaukee Journal Sentinel. You can currently find Emily's work at tonemadison.com and at their own newsletter, Grist From the Mills.
Subscribe to The Recombobulation newsletter here and follow us on Facebook and Instagram at @ therecombobulationarea.
Already subscribe? Get a gift subscription for a friend.
Part of a group who might want to subscribe together? Get a group subscription for 30% off!
 | A guest post by
|