Close your eyes. Imagine this. You just had the most amazing day of your life, work was a breeze, paycheck in hand, and suddenly, you come home, and a bill is at your doorstep. Upon seeing the energy bill, you’re shocked…baffled even that the amount of said bill transcends the amount of your paycheck! As a matter of fact, it is almost 95% higher than it was just five years ago! How could this happen, even when you weren’t the culprit? Well, I’ll tell you what happened: it’s the data centers’ fault. They take our source of water and make our energy bills go crazy high by increasing massive demands on the power grid, forcing expensive shared infrastructure upgrades, and even raising regional energy prices! This is a problem because data centers take up 176 terawatt-hours of electricity, which is a unit of energy representing a trillion watts of power generated and used in an hour, and that’s predicted to double or even triple by 2028 (Lawrence Berkeley National Laboratory). AI data centers in the District of Columbia are drawing natural resources from us, and are sending power bills skyrocketing. Data centers must become less water-dependent in the future.
AI data centers in the District of Columbia are drawing natural resources from us, and are sending power bills that are proliferating. As stated in Bloomberg, “Bloomberg analyzed this data in relation to data center locations, from DC Byte, and found that electricity costs as much as 267% more for a single month than it did five years ago in areas located near significant data center activity.” This evidence highlights the need for justice, as data centers consume a significant amount of electricity. More unfortunately, residents and businesses have to suffer for data centers’ water usage by paying through utility bills for infrastructure upgrades and increasing water rates. It just wouldn’t be fair to pass high industrial energy demands to people who were never involved in it. From the Environmental and Energy Study Institute, a single data center can draw approximately 300,000 gallons of water, while each AI query uses about 0.16 gallons. This means a typical data center contains tens of thousands of servers. Because these servers generate a lot of heat and risk failing or shutting down completely, large amounts of water are required to cool them, placing significant demand on our water sources.
Once the water touches the servers’ systems, it is vaporized, making this method a highly efficient one, yet a dangerous one still, because it uses our water source and turns it to water vapor just to cool off the servers inside. As you can see, data centers cause us to pay for something we don’t even contribute to causing. In response to these issues, the Environmental and Energy Study Institute says that data centers don’t have to use water and potentially save us from our bills, saving us more money than we would have. Based on “AI, Data Centers, and the U.S. Electric Grid: A Watershed Moment, “For example, data centers may be able to improve grid reliability by reducing their power usage during peak periods; however, it is unclear which incentives would best encourage these practices. Theoretical solutions must be translated into effective, real-world policy initiatives that consider economic, political, and social realities as well as technological feasibility.” This evidence strengthens my solution because it will allow data centers to consume less water than they already do. This will also decrease the chance that data centers will consume more water in the future. Based on “Why AI uses so much energy — and what we can do about it”, “One approach is to optimize AI models to use fewer resources without significantly compromising performance, making AI more energy efficient. Instead of training large general-purpose models from scratch, researchers can develop domain-specific AI models that are customized for particular fields, such as computational chemistry or healthcare, reducing the computational overhead.” This evidence strengthens my solution because data centers can be less dependent on just water and more on other sources to lessen water consumption. These solutions work best for preventing data centers from taking our water source up to the point where there is a future with no water.
Now you know why your power bills go up and how to prevent it from happening. In case you forgot, we’re talking about the energy use at AI data centers, which is the cause of our bills always going up, even though we aren’t the ones at fault. Average data centers take around 300,000 gallons of water, while larger ones take around 528,000 gallons of water, according to the Environment and Energy Study Institute. We should try to prevent data centers from destroying our source of water any further and find ways to decrease the amount of water they use. Back to that story…now you’re feeling nothing but rage and want to stop these prices from increasing again and again. Let’s all stop data centers from consuming our money by making them less reliant on our water source!
