While emerging artificial intelligence technology offers exciting possibilities, the long-term impact on the environment is increasingly being discussed. With the introduction of artificial intelligence into our daily lives, data centers have begun to fill up with GPU and CPU hardware with massive computing capacity. However, according to a new study by the Chinese Academy of Sciences, the University of California and Lehman University, the problem of electronic waste (e-waste) from this hardware should not be ignored. The researchers estimate that AI could generate a total of 5 million tons of e-waste between 2020 and 2030, equivalent to the weight of about 25 billion iPhone 16 Pro units!
The unseen cost of AI: 500,000 tons of e-waste a year!
The majority of e-waste originates from data centers in North America, East Asia and Western Europe. According to the study, 58% of this waste comes from North America, while East Asia accounts for 25% and Western Europe 14%. Hardware in data centers is typically replaced every few years as AI technology moves towards applications that require more power and speed.
Large tech companies in particular prefer to buy the latest hardware to stay ahead of the AI race. This leads to the rapid scrapping of old hardware and an increase in e-waste. Researchers suggest a simple solution: If the lifespan of servers in AI data centers is extended, the amount of e-waste can be reduced by up to 58%.
For example, the modules of some AI devices could be reused for less demanding jobs. However, it is a matter of debate whether big companies will welcome this proposal. Because technology giants insist on using the most up-to-date hardware in the rapidly advancing field of artificial intelligence. Companies like Meta are also shelving plans to extend the lifespan of servers, preferring to switch to next-generation devices.
It seems inevitable that this rapid hardware refresh cycle in the AI world will not only increase our carbon footprint, but also increase the e-waste problem. For an eco-friendly AI future, hardware needs to be used more sustainably and recycling processes need to be improved.
In the shadow of advancing technology, e-waste is becoming a bigger environmental problem and the solution lies not only in using technology but also in managing it properly. What do you think about this issue? Please share your views in the comments section below.
{{user}} {{datetime}}
{{text}}