Technology critic Paris Marx argues that America’s growing opposition to AI data centers can draw lessons from international disputes over energy, water and infrastructure while exploring more community-controlled approaches to technology development. (Source: Image by RR)

Communities Seek Greater Oversight of Large Computing Facilities

Growing opposition to artificial intelligence data centers across the United States is following a pattern already seen in countries confronting the enormous energy, water and infrastructure demands of large technology companies. In a wired.com interview, technology critic and author Paris Marx argues that American communities can learn from earlier disputes in Ireland, the Netherlands, Singapore and other countries where rapid data center development prompted restrictions, public protests and questions about resource allocation. Marx, whose new book Hyperscale: The Ambition and Excess of Big Tech’s Data Empires examines the global expansion of computing infrastructure, says the current US backlash reflects concerns that have been developing internationally for years.

Ireland provides one of the clearest examples of how data center expansion can strain electricity infrastructure. Dublin’s grid operator restricted new connections in 2021 after determining that additional facilities could exceed available capacity. According to Marx, some operators responded by exploring projects elsewhere in Ireland or developing dedicated natural gas power generation to supply their facilities. The shift raised questions about fossil fuel emissions, public energy infrastructure and whether private technology companies should gain greater control over electricity systems. Similar concerns have emerged in the United States, where data center developers are increasingly considering dedicated power plants to meet growing demand from AI computing operations.

Water consumption and long-term infrastructure investment present additional challenges. Marx points to Google facilities in The Dalles, Oregon, where a dispute over public disclosure revealed that the company’s operations accounted for approximately 29 percent of the community’s water use during the period examined. He also questions whether the rapid construction of AI-focused data centers and associated power infrastructure could leave communities with expensive or underutilized facilities if projected demand fails to materialize. Drawing comparisons to the excess telecommunications infrastructure built during the dot-com boom, Marx argues that the relatively short useful life of advanced AI chips could make some investments particularly vulnerable to changing technology and market conditions.

Beyond restricting development, Marx highlights alternative approaches that give communities greater control over technological infrastructure. One example comes from New Zealand, where a Māori-led initiative developed speech-recognition technology for the Māori language while using local cloud providers to maintain control over the data. Marx argues that governments should evaluate the social and environmental value of large computing projects alongside their commercial promises, rather than treating continued expansion as inevitable. His broader position is that AI infrastructure decisions should consider not only how much computing capacity companies want to build, but also whether the resulting benefits justify the electricity, water, land and public resources required.

read more at wired.com