ChatGPT's Thirsty Data Centers are Draining Water Resources
ChatGPT's AI chatbot created by OpenAI is becoming increasingly popular, but a recent study reveals that the system's thirsty data centers are consuming large amounts of water resources. Researchers from the University of Colorado Riverside and the University of Texas Arlington examined the environmental impact of AI training and discovered that Microsoft used 185,000 gallons of water when training GPT-3. This is the same amount of water that a nuclear reactor requires for cooling.
The ChatGPT system uses around 500 ml of water for a typical short conversation of 20-50 questions and answers. While this might not seem like much, the cumulative effect of all ChatGPT users can add up to a lot of water usage. Researchers have developed a system to calculate the amount of freshwater required for cooling servers and powering data center servers, and their results indicate a significant drain on water resources.
As the next-generation GPT-4 prepares to launch, the amount of water required by computer systems is likely to increase further. The researchers urge companies to provide more information on how much water they use and what measures they are taking to reduce their water consumption. This data could play a crucial role in addressing the global water crisis.
