China’s commitment to artificial intelligence (AI) is unwavering, with a keen focus on expanding its data centers to accommodate the burgeoning technological landscape. However, this ambitious drive comes at a cost – a substantial amount of water consumption. The recent findings by the Hong Kong-based non-profit organization, China Water Risk, suggest that the country could potentially gulp down a staggering 343 billion gallons of water within its data centers, an amount equivalent to the residential water usage of 26 million people. This revelation sheds light on the environmental impact of AI advancement on water resources.
To prevent data centers from overheating due to the high-powered equipment they host, companies rely on water for cooling purposes. China Water Risk’s report projects a potential tripling of data center numbers by 2030, foreseeing the accommodation of approximately 11 million data center racks. This substantial growth is fueled by the escalating demand for AI services, which inevitably leads to a surge in water consumption on a global scale. Notably, in the United States, researchers discovered that the training of GPT-3, a renowned AI model developed by OpenAI in collaboration with Microsoft, guzzled a staggering 185,000 gallons of water. This excessive water usage is a cause for concern, especially considering its implications for regions already grappling with water scarcity.
The environmental repercussions of AI’s insatiable thirst for water are not confined to a single country. Google’s admission of a whopping 5.6 billion gallons of water usage in 2022 underscores the magnitude of the issue. AI chatbots, such as OpenAI’s ChatGPT, are considerably energy-intensive due to their reliance on specialized chips that devour substantial amounts of electricity. The colossal water consumption associated with these AI technologies poses a significant threat to regions already facing water stress. The exponential growth of data centers and the burgeoning demand for AI services have raised alarms among experts regarding the escalating energy and water consumption trends.
The potential environmental consequences of the water-intensive nature of AI technologies have prompted scrutiny from various quarters. Reports of tech giants attempting to conceal their data centers’ water usage, such as Microsoft’s efforts in Arizona, highlight the growing concerns surrounding this issue. Arm Holdings Plc CEO Rene Haas’ forecast that the world’s data centers could surpass India’s electricity consumption by the end of the decade underscores the urgency of addressing the environmental impact of AI advancement. As AI continues to reshape industries and revolutionize technology, sustainable practices must be prioritized to mitigate the escalating water and energy demands associated with these advancements.
By
