• Tech Tech

Professor calls out little-known aspect of generative AI with major implications for the environment: 'Do you really need these big models that are capable of anything?'

"This is one of the most important ethical implications of artificial intelligence that people are not talking about."

"This is one of the most important ethical implications of artificial intelligence that people are not talking about."

Photo Credit: iStock

We've all heard rumblings about how AI is a menace to humankind (and also bad at making images of fingers), but there's a new environmental angle to its impact. 

A viral TikTok video reveals how power-hungry AI processes can really be.

@professorcasey Environmental impact of AI. #aiethics #sustainability #climatechange #generativeai #huggingface ♬ original sound - Professor Casey Fiesler

The clip, shared and voiced by Professor Casey Fiesler of the University of Colorado, Boulder (@professorcasey), references details from an MIT article about the topic. 

"This is one of the most important ethical implications of artificial intelligence that people are not talking about. Making an image with generative AI uses as much energy as charging your phone," as Fieler declares. 

The study was led by Sasha Luccioni, an AI researcher, and appeared in MIT Technology Review, although it was not peer-reviewed at the time. It detailed how these systems burn through energy, especially when asked to generate images.  

AI tools like the one mentioned have been added to devices in growing numbers, which means processing requests by users can add up quickly. Recent studies found that AI and the chips required to operate it could burn through enough energy by 2027 to power a small country, and that market is projected to grow 42% in the next 10 years. 

As the source states, the researchers looked at the energy that computers consume running these types of requests, while also looking at carbon output related to the entire process. Generating 1,000 images is responsible for the same planet-warming carbon pollution as driving 4.1 miles in a gas-guzzling car. 

Luckily, federal efforts are being funded to innovate new cooling technologies for these data centers, addressing a large portion of the energy use in the sector. Google is also developing a battery-storage project in hopes of using more renewable sources to power them.

Still, some of today's AI processes could be streamlined for specific tasks and be less wasteful, as Luccioni shared and Fiesler reiterates in the video clip.

"If you're doing a specific application, like searching through email, do you really need these big models that are capable of anything? I would say no." 

Some TikTok commenters expressed their own opinions on AI and its power-hungry machinations, with one saying, "I never understood how AI bots are going to be profitable given the cost of energy."

"It's nice the environmental impact is finally being looked at," another shared. "And let's not forget, that's just AI. [Has] anyone studied social media energy usage yet?"

Join our free newsletter for weekly updates on the latest innovations improving our lives and shaping our future, and don't miss this cool list of easy ways to help yourself while helping the planet.

Cool Divider