The water used to cool the AI is a concern for the environment!

During the development and operation of artificial intelligence models, the amount of water used in cooling systems was investigated to prevent equipment from overheating. It is estimated that 4.2-6.6 billion cubic meters of water can be consumed in 2027 with increasing artificial intelligence applications.

Noting that recent research has shown that models such as GPT-3 and BLOOM can lead to significant water and energy consumption, Asst. Prof. İnci Karakaş stated that "This situation is very worrying as the difficulty in accessing potable and usable water is one of the most urgent problems shared by humanity.”

Üsküdar University Vocational School of Health Services (VSHS) Graduate School of Health Sciences Deputy Director Asst. Prof. İnci Karakaş evaluated the water resources and environment used by artificial intelligence applications.

"Their applications have many advantages in terms of environment"

Stating that artificial intelligence applications have many advantages in terms of environment, Asst. Prof. İnci Karakaş expressed that "With artificial intelligence applications, energy consumption can be analyzed and potential savings areas can be determined and the best use of energy can be ensured. In terms of water resources, the need for water is minimized with algorithms that determine water consumption. Limited usable water resources can be protected and irrigation systems specific to the product type can be developed with smart irrigation systems."

"It is stated that models such as GPT-3 and BLOOM can cause serious water and energy consumption"

On the other hand, when the latest research on artificial intelligence is examined, it is stated that various models such as GPT-3 and BLOOM can lead to significant water consumption and energy consumption, Dr. Karakaş continued her remarks as follows:

"In the study conducted at the University of California, it is estimated that 4.2-6.6 billion cubic meters of water might be consumed in 2027 with increasing artificial intelligence applications. This situation is also very worrying as the difficulty in accessing potable and usable water is one of the most urgent problems shared by humanity due to the rapidly increasing population and depleting natural water resources. It will be important for the environment to clearly specify the amount of water used in cooling systems to prevent equipment from overheating during the development and operation of various artificial intelligence models. Along with the water footprint, which shows the amount of fresh water used to produce a product or service, the carbon footprint also needs to be revealed to create environmentally sustainable AI models."

"Potable and usable water resources are limited"

Stating that potable and usable water resources are limited and unevenly distributed around the world, Asst. Prof. İnci Karakaş said that "Water drawn from underground or surface water sources for agricultural or urban purposes, as well as for artificial intelligence applications, can reduce the usable water source. After water is withdrawn from underground and surface water sources, the change in water quality contributes to water stress levels for subsequent use. Global water resources are insufficient in terms of both quantity and quality, and the risk of water scarcity can be faced."

“Water scarcity is already affecting 4 billion people"

Explaining that severe water scarcity can already affect 4 billion people, about two-thirds of the global population, for at least one month every year, Asst. Prof. İnci Karakaş stated that "Studies emphasize that almost half of the world's population may suffer from serious water shortages by 2030 and approximately one out of every four children worldwide will live in regions with serious water shortages by 2040 if integrated and inclusive approaches are not brought to the global water problem. Moreover, the place and time when artificial intelligence applications are trained and run can also be important in terms of water consumption."

"There is not enough awareness"

Stating that there is not enough awareness about the water resources used by artificial intelligence applications, Asst. Prof. İnci Karakaş said that "It can be very difficult to ensure the sustainability of water, as the amount of water consumed during the development and operation of the applications is not clearly revealed. Failure to specify the water footprint could also pose a potential obstacle to environmentally sustainable AI applications in the future. In terms of sustainable artificial intelligence applications, it is important to evaluate the water footprint and carbon footprint together."

"The environmental risks that artificial intelligence applications may pose should be revealed"

Stating that artificial intelligence applications can be used in areas such as environmental science, health, education and research, scientific publishing, digital assistants, transportation and logistics, finance and translation, Asst. Prof. İnci Karakaş concluded her remarks as follows:

"The environmental risks that artificial intelligence applications, which are used in many areas of our lives, should be revealed because AI programs are complex, and they require more energy than other computing systems. It is extremely difficult to determine exactly how much energy artificial intelligence tools consume.

"The effects of global warming may be felt more seriously"

In a study conducted in 2022, it was estimated to emit 24.7 tons of carbon dioxide emissions as a result of training the BLOOM model and 50.5 tons of carbon dioxide emissions as a result of the production of the equipment and the operation of the model. Carbon dioxide is seen to have the largest share in greenhouse gas emissions. With increasing industrialization and population growth, the effects of global warming may be felt more seriously by increasing the temperature in the atmosphere as artificial intelligence applications increase carbon dioxide emissions.

Due to global warming, sea level rise, melting of glaciers, strong winds and heavy rains, deterioration of air quality, decrease in natural water resources, water scarcity and an increase in epidemics may occur."

Üsküdar News Agency (ÜNA)