NEW YORK, NY – SEPTEMBER 22, 2017: Young men use their smartphones as the wait in line for a Uniqlo clothing store to open on Fifth Avenue in New York, New York. (Photo by Robert Alexander/Getty Images)
Getty Images
The United States experienced the warmest summer on record. Yep, it was even warmer than 1936. “The average temperature for meteorological summer (June–August) 2026 was the warmest for the contiguous United States (CONUS) in the 132-year record, exceeding the previous records, set in 1936 and 2021, by 0.4°F,” according to NOAA.” For the planet, August was tied for the hottest month on record. As of September 1, drought covered nearly 60% of the continental U.S. while Hawaii experienced its wettest first half of a year on record and a parade of hurricanes. Our naturally-varying climate system has a “human steroid” sitting on top of it, and some people worry that frivolous AI usage is further amplifying the challenge. Let’s explore the premise.
Environmental Impacts Of AI
To be clear, I use AI. I am not a person that defaults to the perspective that technological advancement is bad, and the “good ole days” were better. In some of my graduate work at Florida State University in the early nineties, I actually explored the use of machine-learning techniques to develop algorithms to track hurricanes. AI can increase productivity, efficiency, and provide breakthroughs beneficial to society. It is also quite useful for mundane tasks like summarizing message threads, organizing thoughts and some aspects of report generation.
Even with these value-added propositions, there is appropriate concern about the ethics of AI, hallucinations, biases, reduction of critical thinking skills, loss of human engagement, and workforce implications. These were among the “bad” in a recent article published in Virginia Tech Engineer. Walid Saad is a computer and electrical engineer at the university. “Unfortunately, AI may have its own carbon footprint and negative environmental impact because it relies heavily on computing at data centers,” he told the writers. “Those data centers consume a large amount of electricity, and also require a significant amount of water for cooling purposes,” He went on to say. This concern increasingly worries me also in the face of climate change, stressed water supply, and intensifying weather extremes.
MIAMI, FLORIDA – JANUARY 14: In an aerial view, high voltage power lines run through a sub-station along the electrical power grid on January 14, 2026 in Miami, Florida. (Photo by Joe Raedle/Getty Images)
Getty Images
Saad noted that sustainable, green infrastructure and AI algorithms are likely on the horizon, but we are not there. Just this week, students in my urban climate class at the University of Georgia read an essay discussing the IPAT equation which was developed by John Holdren, Paul Ehrlich, and Barry Commoner. The equation is simple: I = P × A × T, where I represents our impact on the environment. The other parts of the equation are population (P), affluence (A) and technology (T). The evolution of AI and data centers is at the heart of the IPAT equation. I challenged my students to think about the independent and interconnected aspects of the equation.
“According to a new study from UN University (UNU), AI-related water consumption could equal the basic annual domestic needs of 1.3 billion people by the end of the decade, while its land footprint may exceed 14,500 square kilometres — roughly twice the size of the Jakarta metropolitan area,” according to a recent United Nations press release. “Public debate has largely centred on the energy required to train advanced AI models, but the study finds that day-to-day usage accounts for roughly 80 to 90 per cent of total energy demand,” The report went on to say. They documented that a particularly popular AI service likely produces nearly 2.5 billion prompts every day. That is equivalent to hundreds of gigawatt-hours of electricity annually.
Leave a comment