AI and Water: The Complicated Truth Behind Data Centers
Meanwhile, concerns about AI data centers’ water use have exploded, but experts say the reality is far more complicated than people think.
Journalist Karen Hao recently acknowledged a substantial error in her book Empire of AI, where she claimed a proposed Google data center in Chile could require “more than one thousand times the amount of water consumed by the entire population.”
However, thanks to a correction by Andy Masley, the head of an effective altruism organization, the actual figure appears to be off by a magnitude of 1,000.
The AI Water Issue: Fact or Fiction?
Consequently, Masley has been questioning some of the numbers and rhetoric common in popular media about water use and AI on his Substack.
His main post, titled “The AI Water Issue Is Fake,” has been linked by other writers with large followings, including Matt Yglesias and Noah Smith.
Moreover, experts I spoke to agreed that people often have a muddled understanding of how data centers use water, and that their overall consumption, in many places, is less of a risk than the public may think.
How AI Uses Water: The Reality
Therefore, it’s worth understanding what, exactly, data centers are using water for, and how popular estimates are produced.
Onsite at a data center, water is mostly used for cooling, which involves circulating water through processors to keep them at the right temperature.
In addition, the water that absorbs the heat is then transferred to a cooling tower, where some of it evaporates.
The Complexity of Water Use in Data Centers
Meanwhile, the amount used depends heavily on the individual data center, and using more water means that data centers can avoid running electric cooling systems.
However, using more electricity, by contrast, lessens the water footprint, but ups the power bill—and causes more greenhouse gas emissions.
Therefore, experts say that every location and every state is different, and how much water you will need for the same amount of AI depends on the climate, technology used, and energy mix.
Complicating Factors: Indirect Water Use and Calculations
Consequently, some calculations around AI and water also include indirect water use—mainly from the massive power generation needed for data centers—to estimate their total water footprint.
However, these numbers are generally much bigger than onsite use, but the calculations themselves are region-dependent.
Moreover, experts say that offsite water use from energy shouldn’t factor into data-center water footprints, simply because we don’t tend to count this use when we talk about other industries.
The Need for Nuance in Water Use Discussions
Therefore, it’s worth having a bigger conversation about how and why we’re choosing to use water to cool data centers in the first place.
Meanwhile, experts say that finding out details about water use in a specific data center isn’t always easy, as many companies use nondisclosure agreements to hide even basic information about projects from the public.
However, it’s essential to understand the complexities of water use in data centers to make informed decisions about their environmental impacts.
Source: Link
