How Thirsty is Generative AI?

Behind every chatbot response, AI-generated image, and large language model is a vast network of data centers consuming enormous amounts of electricity and water. In this episode, Emily talks with Dr. Amanda Smith of Project Drawdown about the hidden environmental footprint of generative AI and what it means for a warming, resource-constrained world. Dr. Smith explains how data centers operate, why they are often located where power is cheap and reliable, and why water remains the most efficient way to cool the servers that power today’s AI systems. We unpack the difference between carbon footprints and water footprints, explore why training AI models is especially energy intensive, and clarify common misconceptions about how much water tools like ChatGPT actually use. The key question is not whether we should use AI, but how we use it. Thoughtful deployment, smarter infrastructure, and informed users will shape whether generative AI becomes part of the climate problem or part of the solution.