Convenience for you, Cost for the Climate

By: Kiara Aggarwal, Marissa Monteiro,and Lyzeil Escalo

“Summarize this essay.” “My stomach is hurting, what do I do?”, “Rate my Instagram feed” are some common searches nowadays generated by AI such as CHATGPT. But DID YOU KNOW  that a simple chat of 20-50 questions consumes a quarter of the water that a human body needs to take every single day? That means each question consumes two cups of water. ChatGPT, as a mainstream AI platform, consumes gallons of water at data centres to function accordingly. 

What are data centres? Data centres are the powerhouse of AI models; they generate a substantial carbon footprint while consuming a large amount of energy. A physical building dominated by temperature stores computing infrastructures, notably software, servers, and storage devices. Major companies often rely on it as the bedrock for conveying services online. According to Mclean (April 2023 Earth.org), ChatGPT emits 8.4 tons of carbon dioxide per year, more than twice the amount that an individual emits. This consumption contributes to the decline of freshwater and the rise of climate change. 

There is no doubt that AI models are beneficial and a technological advancement, but the cost of them, if abused, is irreplaceable and irreversible. The demand for ChatGPT nowadays, especially in younger generations, is alarming. We need to be aware of the cost of our decisions. Whether it be to ask an opinion or translate a sentence from a different language, such queries might seem harmless, however, those “small” questions can lead to large consequences in nature. Each interaction consumes energy, water, and adds more to the carbon footprint of AI infrastructure, so let us keep in mind the consequences. Being aware of the cost of AI is the first step to responsible use.

 The convenience of AI comes with a cost, a cost that could be inconvenient for all of us. 

Alternatives for CHATGPT

• Google Gemini

- It consumes 0.6 liters less water (per image) than CHATGPT.

• Perplexity AI

- It consumes 0.005 kWh less (per text prompt) than CHATGPT.

References

McLean, S. (2023, April 28). The Environmental Impact of ChatGPT: A Call for Sustainable Practices In AI Development. Earth.org. https://earth.org/environmental-impact-chatgpt/

Zewe, A. (2025, January 17). Explained: Generative AI’s environmental impact. MIT News; Massachusetts Institute of Technology. https://news.mit.edu/2025/explained-generative-ai-environmental-impact-0117