Welcome to DU! The truly grassroots left-of-center political community where regular people, not algorithms, drive the discussions and set the standards. Join the community: Create a free account Support DU (and get rid of ads!): Become a Star Member Latest Breaking News Editorials & Other Articles General Discussion The DU Lounge All Forums Issue Forums Culture Forums Alliance Forums Region Forums Support Forums Help & Search

OKIsItJustMe

(21,370 posts)
3. Here's an interesting article...
Mon Jun 23, 2025, 07:24 PM
Jun 23
https://www.wired.com/story/ai-carbon-emissions-energy-unknown-mystery-research/
Molly Taft Science Jun 19, 2025 6:00 AM

How Much Energy Does AI Use? The People Who Know Aren’t Saying

A growing body of research attempts to put a number on energy use and AI—even as the companies behind the most popular models keep their carbon emissions a secret.



As a result of this lack of transparency, Luccioni says, the public is being exposed to estimates that make no sense but which are taken as gospel. You may have heard, for instance, that the average ChatGPT request takes 10 times as much energy as the average Google search. Luccioni and her colleagues track down this claim to a public remark that John Hennessy, the chairman of Alphabet, the parent company of Google, made in 2023.

A claim made by a board member from one company (Google) about the product of another company to which he has no relation (OpenAI) is tenuous at best—yet, Luccioni’s analysis finds, this figure has been repeated again and again in press and policy reports. (As I was writing this piece, I got a pitch with this exact statistic.)

“People have taken an off-the-cuff remark and turned it into an actual statistic that’s informing policy and the way people look at these things,” Luccioni says. “The real core issue is that we have no numbers. So even the back-of-the-napkin calculations that people can find, they tend to take them as the gold standard, but that’s not the case.”



Perhaps most crucially for our understanding of AI’s emissions, open source models like the ones Dauner used in his study represent a fraction of the AI models used by consumers today. Training a model and updating deployed models takes a massive amount of energy—figures that many big companies keep secret. It’s unclear, for example, whether the light bulb statistic about ChatGPT from OpenAI’s Altman takes into account all the energy used to train the models powering the chatbot. Without more disclosure, the public is simply missing much of the information needed to start understanding just how much this technology is impacting the planet.

Recommendations

1 members have recommended this reply (displayed in chronological order):

Latest Discussions»Issue Forums»Environment & Energy»Advanced AI models genera...»Reply #3