Popular Artificial Intelligence software such as OpenAi's ChatGpt and Google's Bard consume a lot of energy and require large data centers, cooling these facilities involves a huge waste of water.
For example, a study by the University of Colorado Riverside and the University of Texas at Arlington calculated that training ChatGpt-3 consumed 700,000 liters of fresh water for data center cooling.
A quantity sufficient to make 370 BMW or 320 Tesla cars.
While the exchange of conversations of an average user with ChatGpt is equivalent to the consumption of a bottle of water.
Given the unprecedented popularity of these chatbots, researchers fear that all this water consumption could have a worrying impact on water supplies amid growing drought and environmental concerns.
The researchers also expect that these water consumptions will increase further with newer models, such as Gpt-4, which are based on a larger set and processing of data than predecessor software.
Water consumption issues aren't limited to OpenAI models.
Scholars have also made calculations on Google's water consumption.
Its U.S. data centers collectively drank 12.7 billion liters of fresh water in 2021, about 90 percent of which was drinkable.
"The water footprint of artificial intelligence models can no longer remain under the radar - explain the authors of the study - it must be addressed as a priority and as part of the collective efforts to combat global water challenges".