What is the environmental impact of AI data centers? When text entered into our smartphones or computers is seamlessly completed with the help of artificial intelligence (AI), we often praise the convenience of this technology. But have we ever considered the environmental impact of the countless data centers operating behind the scenes of this AI? A recent report by The Guardian, based on UK government data, is shocking. Estimates suggesting that AI data centers' carbon emissions could reach up to 123 million tons over the next decade are not merely a future warning; they provide an immediate reason to seek solutions. AI is lauded as a technology with the potential to improve human lives in the long term. However, the energy consumption problem, which this technology greedily devours, has reached a level that can no longer be overlooked. Data centers for AI are based on massive server farms, consuming hundreds of megawatts of power and incurring enormous cooling costs due to the accompanying heat emissions. According to The Guardian's report, it is noteworthy that the UK government has underestimated these emissions by more than 100 times. This is not merely a calculation error but suggests a systematic underestimation of the environmental impact of AI infrastructure during the policymaking process. How, then, should we find a balance between the benefits of AI and its environmental costs? The first point to note is the carbon footprint of AI data centers. According to The Guardian's report, the carbon emissions from data centers are comparable to those of large-scale industrial facilities, with some estimates equating them to the emissions of 2.7 million people. This figure is not just a mere list of numbers. 2.7 million people is approximately 80% of Busan's population. The fact that a single piece of infrastructure for AI technology development emits as much carbon as a major city is an issue that cannot be taken lightly in an era of climate crisis. Not only in the UK but also in South Korea, the number of data centers is rapidly increasing. This can be attributed to the expansion of AI technology and cloud demand, and if this trend continues, it could become a significant obstacle to our country's climate change response. South Korea, in particular, as a nation that has formalized its 2050 carbon neutrality goal, faces the dilemma of simultaneously addressing two challenges: fostering the AI industry and achieving climate targets. As the data center industry grows, electricity demand increases exponentially, but the renewable energy infrastructure to support this is not yet sufficient. The second issue to examine is the limitations of energy efficiency. Beyond power consumption, AI data centers require a significant amount of energy for operating cooling systems. Large-scale cooling systems are essential to dissipate the heat generated by high-performance servers running 24/7, leading to additional electricity consumption in this process. Especially in regions like South Korea, with high summer temperatures and humidity, the operation of cooling systems becomes even more critical, exacerbating energy consumption. Technological Advancement is Inevitable, But the Problem Lies in Alternatives In this situation, the development of technologies that dramatically improve energy efficiency can be considered the only breakthrough for AI technology and climate change issues to coexist. The Guardian's report emphasized the urgent need for improved efficiency in AI models and hardware, as well as the decarbonization of energy grids. For instance, some companies have recently attempted to transition to eco-friendly cooling technologies and renewable energy-based facilities, but the pace remains slow. Some advanced companies are adopting natural cooling methods that utilize the cold climates of Northern Europe or attempting innovative approaches like underwater data centers, but these are still in experimental stages or limited in scale. Moving to the third point, there is an attitude that overlooks environmental issues for the sake of technological advancement. Conservative and liberal-leaning media outlets such as The Wall Street Journal and The Economist generally tend to emphasize the importance of technological innovation and economic growth. They argue that the development of advanced technologies, including AI, can improve human lives and, in the long run, contribute to solving environmental problems. Their logic is that technological advancement inevitably entails initial costs and side effects, and thus market-driven solutions are needed rather than regulation. They suggest that industries should be supported in a way that increases efficiency in the long term, rather than being hindered. However, is such an attitude truly realistic? The Guardian presented a contrasting view, emphasizing that both regulation and efficiency improvements must be pursued simultaneously. The fact that the UK government underes
Related Articles