Artificial intelligence (AI) is transforming industries and enhancing our daily lives, but it’s also contributing to a growing concern: electronic waste, or e-waste. A recent study has warned that the rise of generative AI technology could lead to a dramatic increase in e-waste by 2030. This article breaks down the potential environmental impact and offers insights on how sustainable practices can help mitigate the effects.
With AI advancing rapidly, tech companies are constantly upgrading their data centres with new hardware, servers, and specialised AI processors. This relentless cycle of hardware upgrades means that older devices are frequently discarded, resulting in a significant rise in electronic waste. According to the study, the growth of AI may increase e-waste to nearly 2.5 million tonnes by the next decade.
That’s the equivalent of every person on Earth discarding almost two iPhones each.
As generative AI systems continue to grow, so does the demand for high-performance hardware. AI models require immense computing power to process and analyse data, which leads companies to replace servers and GPUs (Graphics Processing Units) every few years. These rapid upgrades, especially with the latest AI technology, quickly render older devices obsolete, piling up e-waste faster than ever.
This e-waste includes materials that are hazardous to human health and the environment, like lead and chromium, as well as valuable metals such as gold and silver. Unfortunately, much of this e-waste is not recycled properly, causing pollution and wasting precious resources.
Read More: Optimizing AI for a Greener Future
In 2022, global e-waste reached a record 62 million tonnes, according to the United Nations, and it’s projected to hit 82 million tonnes by 2030. AI technology could contribute heavily to this surge unless sustainable measures are taken. The study found that around 58% of AI-related e-waste is expected to come from North America, with significant contributions from East Asia and Europe.
Interestingly, geopolitical tensions, especially restrictions on advanced GPUs between the U.S. and China, may contribute to even more e-waste. When data centres in countries like China lack access to the latest AI hardware, they rely on outdated servers, which means more servers are needed to achieve similar performance. This inefficiency leads to even more e-waste, as old servers are discarded more frequently.
Read More: Global Tech Leaders on U.S.-China Tech Rivalry
To curb the e-waste issue, the research suggests a “circular economy” approach. This involves extending the lifespan of AI hardware, reusing older components like GPUs and CPUs, and recycling valuable materials. By prolonging the lifespan of AI servers and reusing components, we could reduce e-waste by as much as 86% globally.
Some strategies to reduce e-waste include:
Extending Hardware Lifespan: Keeping AI servers and hardware in use for longer periods can reduce turnover.
Component Reuse: Repurposing obsolete GPUs, CPUs, and batteries for other applications can cut down on waste.
Efficient Algorithms: Developing AI algorithms that are more efficient can lessen the demand for high-performance, high-waste hardware.
Read More: AI and the Circular Economy: Accelerating the Transition
AI has incredible potential to reshape the future, but we must address the environmental impact that comes with its hardware demands. By promoting sustainable AI practices, we can work towards a balance that allows for technological growth without compromising the planet. As AI technology continues to develop, adopting sustainable measures now will ensure that AI’s positive impacts extend beyond innovation to include environmental responsibility.
Through collaboration between tech companies and governments, we can ensure AI contributes to a better future without piling up e-waste.
By clicking "Accept", you agree to the storing of cookies on your device to enhance site navigation, analyze site usage and assist in improving your experience.
