The development of artificial intelligence (AI) has been a boon to humanity, but it comes at a cost. One of the most significant environmental impacts of AI is the amount of water consumed by data centers that train large language models like ChatGPT. According to a recent report, Microsoft and Google have seen significant spikes in water consumption due to the use of AI technology 12.
Microsoft and OpenAI used a supercomputer in Iowa to train GPT-4, one of the most advanced AI models in the world. The supercomputer required a lot of water to cool down, especially in the summer months 3. A researcher estimates that ChatGPT uses about 500 milliliters of water for every 5 to 50 prompts or questions. The researcher suggests that AI developers should choose locations with lower water and energy consumption for their training 3.
The environmental costs of AI are significant, but Microsoft and OpenAI have acknowledged them and are working on ways to make their systems more efficient and sustainable. They have set some goals to reduce their carbon footprint and water usage by 2030 32. However, it is essential to ask ourselves some questions about sustainability. How can we balance the benefits of AI with its environmental impact? What can we do to reduce the amount of water consumed by data centers? How can we ensure that AI development is sustainable in the long run? These are some of the questions that we need to answer as we move forward with this technology.
