Image: University of Kansas
Kansas: A new study has revealed that artificial intelligence systems used for writing and image creation emit hundreds of times less carbon compared to the same tasks done by humans.
In a joint research study by the University of Kansas and the University of California-Irvine, artificial intelligence systems such as GPT-3, BLOMAI, DALI-2, and others were compared to human performance in writing and image tasks.
Like cryptocurrency, artificial intelligence is also under discussion due to its energy consumption and its contribution to climate change. The long-term impact of human expenditure and environmental effects has been examined, but the comparison between the two is quite moderate.
However, researchers found that the use of artificial intelligence systems for writing results in 130 to 1500 times less carbon dioxide emissions compared to human writing, while image creation systems result in 310 to 2900 times less carbon dioxide emissions compared to human creation.
This study, published in the journal Nature, was jointly conducted by Andrew Torrance from the University of Kansas and Bill Tomlinson, Rebecca Black, and Donald Peterson from the University of California-Irvine.