Data centers — more than 4,000 in the United States — are huge complexes filled with computers and data that power artificial intelligence. These centers need a lot of electricity for power and a lot of water to cool the computers.

On Wednesday, April 22, students gathered in the Pearse Hub for Innovation (PHI) for a conversation on AI’s environmental impact. The session, one of many Earth Week events at Loomis Chaffee, was led by Matt Johnson, director of educational and administrative AI initiatives; science teacher Ned Heckman; and Sarah Griggs, associate director of the Alvord Center for Global & Environmental Studies.

While this event centered on AI’s environmental impact, the discussion also touched on many parts of people’s daily lives that require the use of energy. This was, after all, an Earth Week event, and AI is not alone in draining power. The program helped to put the numbers into context and considered ways we all might be able to reduce our carbon footprint.  

“What we are not trying to do is convince you of a particular narrative,” Ned said. “My guess is some of you have your own thoughts and uses of AI. ... We are not here to shame you on your uses or suggest we have all the answers.”

Before the teachers started presenting, students were asked to share their concerns about AI’s environmental impact. Among the answers were: “High water use.” “Fossil fuel use.” “The mining needed to get rare earth materials.” “I’m worried I’m responsible for doing harm.”

Students were introduced to an interactive tool designed by data scientist Hannah Ritchie that showed the watt-hours (wh) used in various endeavors, from running a gas lawnmower for an hour (9,000 wh) to boiling a kettle of water (100 wh) or running a ChatGPT median query (0.3 wh). It all needs context. How often is that kettle of water boiled? Those ChatGPT queries add up when one considers how many of these queries each person submits in a day, a week, a year. Ned also told the students about an Environmental Protection Agency Greenhouse Gas Equivalencies Calculator, another tool that helps with assessing environmental impact.

“Take a look at what things you do that use the most energy,” Sarah said to the students. She encouraged them to consider where they can move the needle with their own actions.

“By making a little adjustment in our own lives, where can we make the most impact?” she wondered. “And if we all do that, that will help us move forward as a planet together.”

In the 2026–27 academic year, Matt and Sarah will co-teach a course called AI and Climate: Promise and Price. The course description says it “examines AI as both an environmental challenge and a potential solution. We'll begin by establishing a shared understanding of what we mean when we talk about AI, as well as defining climate change and climate science. ... The course pushes students beyond tech optimism and doomerism to understand complex tradeoffs in our response to climate change.”