AI could cause e-waste equal to 10B iPhones by 2030
The rapid evolution and growing computational requirements of artificial intelligence (AI) models, could lead to the creation of electronic waste on the scale of more than 10 billion iPhones per year by 2030. Cambridge University and Chinese Academy of Sciences researchers made the projection, in a study published in the journal Nature. The study seeks to give an early estimate of this looming challenge's scale and potential solutions.
Study explores potential e-waste from AI servers
The researchers' study centered on the possible e-waste from AI servers. They wrote, "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." The researchers also looked at different growth scenarios (low, medium, and high) and their computing resource requirements, and lifespan projections.
AI-driven e-waste could increase thousandfold by 2030
The researchers' findings indicate that e-waste could increase as much as a thousandfold over 2023. They predict e-waste will surge from 2.6 kilotons (kt) per year in 2023 to some 0.4-2.5 million tons (Mt) per year by 2030. This massive increase is due to the rapid growth and changing computational requirements of AI models, which need a lot of hardware that ultimately goes obsolete and adds to e-waste.
Researchers propose strategies to mitigate e-waste
The study also proposed broad strategies to mitigate the projected increase in e-waste. These include downcycling servers at the end of their lifespan, repurposing components related to communications and power, and improving software efficiency. The researchers suggest that these measures could potentially reduce the waste load by 16-86%. However, the effectiveness of these strategies largely depends on their adoption rate and extent of implementation across the industry.
