Using Generative AI a lot? Here's how much energy it takes, according to an MIT researcher.
We can create just about anything in a matter of seconds thanks to generative artificial intelligence (Gen AI). It's fun to generate a meme or throw together a quick word prompt. But perhaps a lesser-known part of this equation is that these tasks use up a lot of energy and environmental resources.
Researchers at the MIT Climate and Sustainability Consortium are studying the direct and indirect impacts of using generative AI. Impacts that show up in almost hidden ways to the end user with added stress on the power grid, along with an environmental tax.
10 times more power
"AI data centers are much more power intensive," said Dr. Noman Bashir, the Computing and Climate Impact Fellow at the consortium. "So, if you have a normal data center, an AI data center would be up to 10 times more power intensive. It would have more power demand."
Think of it this way: One simple web search takes one unit of energy. Doing that same search, but using an AI model instead, takes 10 units of energy. And finally, using AI to generate an image? That will take you up to 1,000 units. Which leads us to the extra environmental tax.
Water needed to cool chips
"There are also impacts of water usage, because these chips that run these models get very hot. And you need water to cool them down," said Bashir.
WBZ-TV wanted to find out how big can one individual person's impact really be. Can we just relax and use AI? Dr. Bashir thinks the answer to that question lies somewhere in between. He says we shouldn't necessarily worry about using AI, especially if it's used responsibly.
"Judicious use of resources"
"Putting everything on the end user and making them feel guilty is not the right approach," said Bashir. "So, I don't think the end user should feel completely responsible to do these and to manage this problem. However, there is a case to be made for judicious use of resources."
Dr. Bashir says while we can certainly limit our usage of the more powerful AI tools, it's ultimately going to take the companies training new models to drive long-term sustainability. Something we can do to help limit the stress on the power grid and the environment is only using generative AI when absolutely necessary.