The results are, frankly, staggering. “Our study aims not to precisely forecast the quantity of AI servers and their associated e-waste, but rather to provide initial gross estimates that highlight the potential scales of the forthcoming challenge,” they wrote. After modeling three different scenarios based on low, medium and high rates of AI growth, the team estimated that AI’s e-waste could grow from 2.6 thousand tons per year in 2023 to 0.4-2.5 million tons per year in 2030.
How much e-waste is GenAI expected to create by 2030?
Answer: The equivalent of 10 billion iPhones.
Generative AI’s carbon footprint is growing as fast as its popularity. A team of researchers from Cambridge University and the Chinese Academy of Sciences recently published a paper in the journal Nature in which they attempt to forecast how much e-waste AI models could be generating by the end of the decade.
The results are, frankly, staggering. “Our study aims not to precisely forecast the quantity of AI servers and their associated e-waste, but rather to provide initial gross estimates that highlight the potential scales of the forthcoming challenge,” they wrote. After modeling three different scenarios based on low, medium and high rates of AI growth, the team estimated that AI’s e-waste could grow from 2.6 thousand tons per year in 2023 to 0.4-2.5 million tons per year in 2030.
Fortunately, they pointed out that adopting greener practices in computing could help keep this number down considerably if everyone puts in the effort. This includes things like downcycling components that reach the end of their lifespan, rather than just tossing them. And improvements in software and efficiency could also extend the life cycle of hardware components.
The results are, frankly, staggering. “Our study aims not to precisely forecast the quantity of AI servers and their associated e-waste, but rather to provide initial gross estimates that highlight the potential scales of the forthcoming challenge,” they wrote. After modeling three different scenarios based on low, medium and high rates of AI growth, the team estimated that AI’s e-waste could grow from 2.6 thousand tons per year in 2023 to 0.4-2.5 million tons per year in 2030.
