Wait... AI uses water too?
Summary
An infographic explaining the hidden water consumption of AI systems due to heat and cooling, highlighting an often-overlooked environmental cost.
Similar Articles
How much water does AI really use?
An exploration of the water consumption associated with training and running AI models, often overlooked in discussions of AI's environmental footprint.
How much of a problem is AI’s water use?
The article explores the growing water consumption by AI data centers, its environmental implications, and potential solutions through technology and strategic siting.
Is AI water crisis actually a hard requirement, or just how we build?
The article examines the water crisis driven by AI data centers, noting that cooling concentrates thermal load in single watersheds, and proposes distributed compute over idle hardware as a potential solution to spread water and energy demand.
The Real Water Cost of AI Isn't Inside Data Centers. But Tech Companies Don’t Report It.
This in-depth investigation uncovers the true water footprint of AI data centers: the bulk of water consumption occurs during electricity generation, not inside the data centers themselves. Indirect water use can be 10 to 24 times greater than direct use. Tech companies largely omit these figures, and residents and farmers in the arid U.S. Southwest are bearing the cost of the AI boom.
AI water usage per 100 tokens
A user inquires about the average water consumption per 100 tokens for frontier AI models, highlighting the environmental impact of AI.