With the rapid advancement of generative artificial intelligence technology, the amount of electronic waste is expected to significantly increase in the coming years. According to a global research analysis, by 2030, AI-related electronic waste is projected to surge from 2,600 tons in 2023 to 2.5 million tons. This figure is equivalent to nearly two iPhones being discarded by each of the world's 8.5 billion people, raising serious concerns about environmental impact.

Robot AI

Image source: Picture generated by AI, provided by Midjourney

The rapid growth of generative AI is forcing frequent updates in hardware and chip technology, leading to a rapid obsolescence of electronic devices. These discarded devices often contain toxic metals such as lead and chromium, which not only pose health risks but can also cause severe environmental pollution. Additionally, many old devices contain precious metals like gold, silver, and platinum, which can be recycled, but the surge in electronic waste makes recycling more challenging.

The research team, from the Chinese Academy of Sciences and Israel's Reichman University, noted in an article published in the journal Nature Computational Science on October 28 that the total amount of electronic waste could accumulate to between 1.2 million and 5 million tons between 2020 and 2030. They mentioned that geopolitical influences, particularly restrictions on semiconductor imports, and the trend of quickly replacing servers to reduce operating costs, could exacerbate this issue.

The study also found that North America (including the United States and Canada) will bear more than half of the untreated electronic waste, estimated at 58%. East Asia (including China, South Korea, and Japan) will contribute 25%, while the EU and the UK will account for 14%. The US restrictions on the sale of high-end GPUs to China will also impact the environment, forcing Chinese data centers to use outdated server models, which not only reduces computational efficiency but also increases the demand for physical servers.

Furthermore, the research team proposed solutions to the electronic waste problem. They recommended implementing circular economy strategies to reduce electronic waste by 86%. Specific measures include extending the lifespan of AI-related hardware, reusing outdated GPUs, CPUs, and batteries, developing more efficient computational algorithms, and improving chip computational efficiency.

Key Points

🌱 By 2030, AI-related electronic waste is expected to reach 2.5 million tons, equivalent to nearly two iPhones discarded per person.

💻 Frequent hardware updates render existing devices obsolete quickly, resulting in a significant amount of toxic electronic waste.

♻️ Implementing circular economy strategies can reduce electronic waste by 86%, highlighting the urgent need for environmental protection and resource recycling.