Thirsty AI? Study says that ChatGPT drink 500ml of water per conversation

According to the research – called 'Making AI less “thirsty”: Discovering and addressing the secret water footprint of AI models' – the training of GPT-3 No. data center state-of-the-art Microsoft in the United States can directly consume 700.000 liters of clean fresh water (enough to produce 370 BMW cars or 320 electric vehicles Tesla) and water consumption would have been tripled if the training were done in data centers da Microsoft in Asia, but this information was kept secret.

According to the study, to respond to global water challenges, AI models can, and must, take social responsibility and lead by example, addressing their own water footprint.

Read also

As the AI ​​industry grows, what impact will it have on the environment?

Artificial intelligence (AI) programs can seem disembodied. But they are powered by networks of servers in data centers around the world, which require large amounts of energy and large volumes of water. Because AI programs are so complex, they require more energy than other forms of computing. But the problem is: it's incredibly difficult to define exactly how much.