A study by researchers at the Institute of Urban Environment, Chinese Academy of Sciences, and their colleagues at Reichman University, Israel, suggests that the use of generative artificial intelligence could lead to the generation of huge amounts of waste.
The forecast is based on component volume calculations, including motherboards and batteries, which will presumably be used to power AI applications by 2030. They run on hardware that will eventually need to be replaced with new ones.
Image source: Pinterest
Researchers worry that AI consumes a lot of resources, which requires a lot of batteries. Each piece of equipment has a limited lifespan. Once that lifespan is over, there are problems with its disposal. To estimate the expected volume of waste, an international group of environmentalists took the amount of equipment installed in an average data center as a calculation. If current rates continue, AI could generate up to 5,000,000 tons of waste in 6 years. This figure could be higher.
https://overclockers.ru/blog/Astra_news/show/188359/Ekologi-prognozirujut-cherez-6-let-II-mozhet-privesti-k-millionam-tonn-elektronnyh-othodov