The Paradox of AI Energy Use w/ Dr. Amanda Turnbull-McRae: 10 June 2026

Last week, the United Nations released a report on the environmental cost of artificial intelligence, giving estimates for the amount of carbon, water, and land that will likely be dedicated to powering AI in the near future. For example, the report predicts that by 2030, Artificial Intelligence could consume 3% of the world’s electricity and require more water than the total amount drunk by the entire population of Earth annually. 

However, experts in AI also predict that the technology will continue to become more efficient as time goes on, requiring less electricity, water and land to operate. To explain how all of this can be true at the same time, some people in the AI world have invoked the so-called “Jevons Paradox”—referring to an insight from the Victorian-period economist William Stanely Jevons, who noticed that the increased efficiency of steam engines went hand-in-hand with increased coal consumption.

This morning, producer Toby spoke to Dr. Amanda Turnbull-McRae, senior lecturer in Law at the University of Waikato, about this apparent economic paradox as it relates to the future of AI.