In recent years, artificial intelligence (AI) has become an indispensable tool in many industries and areas of life, from healthcare, finance, to manufacturing and entertainment.

However, behind the power and potential of AI, there is an increasingly concerning issue: the consumption of resources, especially water, in the operation of AI systems.

AI and Water Pioneering Solutions for a Thirsty Planet.jpg
Using AI without impacting the environment is an urgent problem. Photo: Trendsresearch

One of the key factors that leads to water consumption is cooling data centers. AI systems require a lot of computation, which leads to high heat generation in servers. To keep these servers running efficiently and not overheating, water is used for cooling.

A recent study conducted by The Washington Post in collaboration with researchers at the University of California found that each 100-word email generated by OpenAI's GPT-4 consumes 519 ml of water to cool the system, equivalent to 27 liters in a year if used once a week.

Currently, the US has about 16 million workers. Assuming that 1 in 10 people use AI at the above frequency, the amount of water needed for cooling in a year is 435,235,476 liters, equivalent to the amount of water used by all Rhode Island residents in 1.5 days.

As demand for AI technologies increases, pressure on water resources could become more severe, especially as climate change impacts global water supplies.

Experts have called on tech companies to focus on more sustainable resource use, from optimizing cooling processes to finding renewable energy sources and other water-saving solutions.

Ensuring that AI development does not add to the burden on the environment and natural resources is an urgent challenge that needs to be addressed.

(According to Washington Post)

China's Alibaba 'magically' transforms more than 100 open-source AI models in just one day On September 19, Alibaba announced more than 100 open-source AI models in an effort to improve its competitiveness with technology rivals.