The Advancement Of Data Centers
Artificial Intelligence (AI) has been on the horizon of human invention for quite some time. From the compass, to the internal combustion engine, to the computer, to AI, human technology continues to improve and advance. From movies about evil sentient machines to comics about the human race being controlled by robots, the concept of AI has been a popular trope in the media for decades.
But something that not many saw coming, along with its harsh environmental consequences, were AI data centers. Data centers were originally created by the U.S. government to house mainframe computers (large units used to contain, run, and manage large amounts of data). At the time, these computers were large and incredibly slow, leading to innovation and growth of computers. Despite their many innovations, data centers have always served the same purpose—running and delivering IT infrastructure while housing and managing data. However, these centers have changed with the implementation of AI, generative AI, OpenSource, and other applications. Now, towns shudder at the thought of data centers being built near them.
“Building an AI data center on Vashon would be really bad. It would take up a lot of water, it’s expensive, we don’t have sufficient resources, and they’re super loud,” sophomore Caleb Ressler said. The truth is, AI data centers have become controversial not only due to them acting as the pumping heart of AI software, but due to the enormous amount of water required to cool the hardware.
“Because nearly all of the electricity consumed by a data center ends up as heat, servers and other equipment require continuous cooling. These cooling systems remove the heat generated by computing and release it into the surrounding environment as waste heat,” Karen Seto, the Frederick C. Hixon Professor of Geography and Urbanization Science and Director of the Hixon Center for Urban Sustainability said in an article from Yale School of the Environment titled Data Centers, Urban Heat, and AI Growth. Data center’s constant cooling can lead to temperatures increasing by up to 16 degrees Fahrenheit where they are built. Ironically, this could lead to inhabitants spending more money and energy on cooling their own homes.
“AI data centers have been known to decrease air quality in towns they are built in. That also leads to public discourse and unhappy civilians,” sophomore Nico Youngs said. The air quality issue in areas near centers is real. Noxious gases and particulate matter is emitted into the air from the centers and can cause irritation in the eyes and lungs. This can lead to higher rates of asthma, other respiratory-related issues, and in rare cases, death. However rare this is, data centers still pose a threat to a basic human right—clean air– says a study from Alabama University.
“It would not be good to build a data center near the island. It would cause loads of disturbances and mess with the environment”, sophomore Parker Ward said. One of the main reasons that Vashon will most likely never be the host of a center is due to its lack of pure freshwater. Although Vashon is surrounded by miles of water, an incredibly small amount of it is readily available to be used in the cooling process. The amount of water required to prevent the overheating of centers is astonishing. Overheating has become a prominent issue with AI, especially because the technology is so new and requires vast amounts of energy to run. To deal with this issue, massive cooling systems were installed to not only cool off processor chips, but also to funnel hot air out of the centers. In the pursuit of temperature control, data centers can consume up to 5 millions gallons of water a day. Data centers are tapping into water supply in a time when droughts and water shortages are increasingly common. Every prompt given to AI, each generated image of a kitten with dragon wings, every single tiny lick of AI uses more energy at centers, creates more heat, and therefore requires cooling, further taxing the already diminishing drinkable water supply of our planet. According to statistics from the Environmental and Energy Study Institute (EESI), “only 5% of earth’s water is fresh, and only 0.3% of earth’s total water is drinkable.” In a time when drinkable water is not guaranteed for all, this fact is hard to stomach. “Despite gains since 2015, 1 in 4 – or 2.1 billion people globally – still lack access to safely managed drinking water*, including 106 million who drink directly from untreated surface sources”, a statistic from the World Health Organization (WHO) details. Another issue that has been gaining attention in the field of AI has been how centers are seemingly targeting areas of lower socioeconomic status.
“Data centers are increasingly being sited in low-income neighborhoods and communities of color, areas already burdened by environmental and economic injustice”, Emily Peterson-Cassin, Director of Competition and Market Fairness said. It has become noticeably common for ai data centers to target areas lower on the socioeconomic scale. Typically job hungry, minority dense, educationally disadvantaged, and underserved communities are the types of places being targeted. A recent statistic from Pew Research Center explains that, “67% of planned data centers are in rural areas, while 87% of existing data centers are in urban ones”. It’s interesting that with the advancements of AI, its integration into data centers, and its harmful environmental effects, more and more server rooms are being planned in rural areas. Recently large minority communities in Memphis, Missouri, and Alabama have begun pushing back against an onslaught of centers being built near them. This is not happening in wealthy, suburban, and even urban areas with large white populations. At the same time that systemic issues targeting the disenfranchised become increasingly common, public pushback, rallying, and protesting is as well. Like recent civilians who gathered to protest against data centers on July 18th in Vicksburg, Mississippi, the public will not stand idly by as centers continue to target the underserved.
Recently, AI has gone from the tech of the distant future, to the tech of today. And its transition has not gone unnoticed by the people. Data centers threaten the safety of the public, consume drinkable water like a vampire does blood, and affirms systemic violence towards ordinary people. It’s unsure what the future holds for AI, however what can be said for sure is that the people aren’t happy.
Comments