As artificial intelligence (AI) continues its rapid advancement, we must confront its environmental footprint. Beyond energy consumption, AI’s demand for water is a critical concern that demands attention.
AI data centres require cooling, and substantial amounts of fresh water are utilized for this purpose. Additionally, water is consumed in the mining and production of AI-related hardware.
The production of AI hardware utilizes and contaminates water due to the mining of rare materials like silicon, germanium, gallium, boron, lithium, gold, copper and phosphorus. The extraction of these minerals is resource-intensive and significantly affects the environment, leading to water pollution. AI is hungry for power too – the amount of additional energy required to run an AI data centre is far greater.
AI water footprint
The water consumption associated with AI is not limited to its operational phase. AI hardware production requires intensive mining for rare materials such as silicon, germanium, gallium, boron, and phosphorus. The extraction of these minerals has a substantial environmental impact and contributes to water pollution.
Additionally, AI systems’ water use extends to the production of related hardware. The extensive deployment of AI compounds this problem, putting pressure on the drinking water supplies of local communities. With the increasing adoption of AI, there is additional strain on municipalities, agriculture, and various industries that are already facing challenges due to drought and scarce water resources.
Solutions?
Researchers have explored AI’s potential to address water challenges.
Water Efficiency
AI can optimise water usage in agriculture by automating irrigation systems. Smart machines, robots, and sensors analyse data to enhance efficiency and detect leaks, as demonstrated in smart irrigation systems for crops.
Water quality monitoring
AI-powered biosensors can accurately detect toxic chemicals in drinking water, surpassing current quality monitoring practices.
Sustainable management – a double-edged sword
AI can aid in managing water security and wastewater treatment, contributing to sustainable water use. However, on the practical side, it is the willingness to implement these grand ideas and systems that is quite often the problem.
Uncharted territory
While there is peer-reviewed analysis on the role of AI in managing water and achieving Sustainable Development Goals (SDGs), the direct and indirect effects of AI on water usage are largely uncharted.
Recent reports indicate that AI systems, such as ChatGPT, may use approximately 500 millilitres of water for every 5 to 50 prompts or questions posed by a user. Given the rapid increase in chatbot usage globally, this could lead to a considerable water footprint, raising concerns among experts and environmentalists.
Conclusion
As technological advancement continues, it is imperative to move forward with caution. Achieving a balance between innovation and the responsible use of resources is essential for a sustainable future where AI contributes positively without jeopardizing our essential water resources.