Quick Buzz Feed
New poll finds over half of Americans fear AI-driven data centers will hike electricity costs and strain water supplies.
A recent national survey conducted by The Associated Press, NORC Center for Public Affairs Research, and the University of Chicago's Energy Policy Institute reveals a significant increase in American adults' apprehension regarding the environmental footprint of artificial intelligence. The poll, which surveyed 3,424 US adults in August, found that 53 percent expressed being "extremely" or "very" concerned about AI's ecological effects. This marks a notable 12 percentage point jump from the 41 percent recorded in the previous year's survey, indicating a rapid rise in public awareness and unease about this issue. The concern about AI's carbon footprint now even surpasses worries about traditionally high-emission industries such as meat production and aviation.
The increasing public anxiety is particularly concentrated on the local impact of data centers, which are the backbone for powering AI and cloud computing. The survey highlights that a majority of Americans are profoundly concerned about how these facilities could strain local electricity grids and deplete water supplies in the communities where they are constructed. This concern is not limited to one political group, although Democrats show a higher level of apprehension (two-thirds) compared to independents (half) and Republicans (four in ten). This suggests a broadening understanding across the political spectrum of the tangible, localized environmental consequences associated with AI infrastructure development.
Michael Greenstone, the director of the University of Chicago's Energy Policy Institute, characterized the current rapid expansion of AI as a "massive experiment in real time." He noted that there might be "no putting the milk back in the bottle," suggesting that the changes brought about by AI's growth are irreversible and carry inherent risks. Greenstone attributes some of the public's heightened concern to the natural human apprehension towards significant, fast-paced technological shifts, framing it as a reflection of the "scary" nature of such profound societal and environmental transformations.
Despite some bipartisan legislative pushback and community concerns, the United States is witnessing a vast expansion of data centers, with thousands already operational and many more under construction to meet the surging demand for AI processing and cloud services. This expansion includes both older, conventional facilities and newly developed "hyperscale" centers specifically designed for the intensive resource requirements of modern AI. States like Virginia and Texas are currently leading in data center operations, with California, Ohio, and New York rapidly emerging as key digital infrastructure hubs. Notably, former President Donald Trump has publicly encouraged communities to welcome these tech facilities, emphasizing their potential to create jobs.
A significant concern raised by individuals like Gedeon Kebede, a computer science student from Ohio, is the substantial water usage by AI data centers. Kebede articulates worries that this consumption diverts vital water resources away from other critical uses, such as agriculture and the needs of local farmers and residents. Cornell University energy systems professor Fengqi You further clarifies that while data centers' overall water consumption might be "nationally modest," their localized impact can be "decisive," especially in areas already facing water stress. Nancy Antunes, a retired Democrat from Massachusetts, echoed these sentiments, stating her belief that decisions regarding data center placement prioritize corporate interests over the well-being of the affected communities.