Study looks at reducing carbon footprint of AI

When most people think of carbon emissions, they think of electricity and heat as well as transportation and manufacturing, but scientists are grappling with carbon emissions from another source.

Artificial intelligence can generate volumes of carbon emissions and one study said that one system processed 1,400 pounds of emissions, according to a press release from Stanford University. 

How much is 1400 pounds of emissions? The study said it's the equivalent of a passenger flying round trip between New York and San Francisco. But that's not all.

To generate an AI language system from scratch, up to 78,000 pounds of carbon emissions are used.

The study is looking at ways to decrease its carbon footprint, with a movement in the works to make AI more green and cleaner.

Stanford teamed up with Facebook and McGill University to develop a tool that shows how much electricity a machine uses and how much that means in carbon emissions. 

“As machine learning systems become more ubiquitous and more resource intensive, they have the potential to significantly contribute to carbon emissions,” Peter Henderson, a PhD student at Stanford in computer science and the lead author, said in the press release. “But you can’t solve a problem if you can’t measure it. Our system can help researchers and industry engineers understand how carbon-efficient their work is and perhaps prompt ideas about how to reduce their carbon footprint.”