Your AI tools run on fracked gas and bulldozed Texas land

The rise of artificial intelligence is unexpectedly breathing new life into the fracking industry, a sector previously criticized by environmental advocates for its detrimental effects on water quality and increased seismic activity. AI firms are now establishing vast data centers in proximity to major gas extraction sites, often relying on fossil fuels for their power needs. This trend, often overshadowed by discussions about AI's potential to revolutionize healthcare and address climate challenges, raises critical concerns for the communities hosting these facilities. A recent report from the Wall Street Journal highlights the case of Poolside, an AI coding assistant startup constructing a sprawling data center complex on over 500 acres in West Texas—an area roughly 300 miles from Dallas, comparable in size to two-thirds of Central Park. This facility is set to generate its own electricity using natural gas sourced from the Permian Basin, America's most productive oil and gas region, where hydraulic fracturing is prevalent. Named Horizon, this project will deliver two gigawatts of computing power, equivalent to the electrical output of the Hoover Dam, but sourced from burning fracked gas instead of renewable resources. Collaborating with CoreWeave, a cloud computing enterprise providing access to Nvidia AI chips, Poolside's initiative is part of a wider movement among major AI companies. OpenAI's CEO Sam Altman recently visited the Stargate data center in Abilene, Texas, which is located about 200 miles from the Permian Basin. During his visit, he acknowledged the reliance on natural gas for powering the facility, which demands around 900 megawatts of electricity across eight buildings, supplemented by a new gas-fired power plant. Local residents have expressed concerns over the impact of these developments. Arlene Mendler, a long-time resident near the Stargate center, lamented the loss of tranquility and natural beauty in her neighborhood due to construction. With West Texas grappling with drought, the implications on water resources are particularly alarming. Although Oracle claims each building will consume only 12,000 gallons of water annually after an initial fill-up, experts like Shaolei Ren from UC Riverside warn that the energy demands of these facilities could lead to higher indirect water consumption at the power plants. Meta is also venturing into this territory, planning a massive $10 billion data center in Louisiana, which will require substantial natural gas power generation. Local residents, similarly to those in Texas, have voiced their discontent over continuous construction activities. However, Meta's new facility in El Paso, Texas, is positioned to use 100% renewable energy, marking a slight shift towards cleaner operational practices. The AI sector's pursuit of fracked gas is not merely a matter of energy sourcing but is intertwined with global competitiveness. Industry leaders argue that to maintain a technological edge over countries like China, the U.S. must bolster its energy production capabilities. This sentiment was echoed by Chris Lehane, OpenAI's VP of global affairs, who emphasized the potential for re-industrializing economically challenged areas through such developments. Government support has further accelerated this trend, with policies aimed at streamlining permits for gas-powered AI data centers. Yet, the environmental ramifications of this rapid expansion remain a pressing concern. Many AI users are unaware of the carbon footprint associated with their technologies, focusing instead on innovations like OpenAI's Sora 2 video-generation tool, which demands significant energy resources. As the AI industry continues to intertwine with fracking, it raises profound questions about sustainability and the future of energy consumption. A recent study from Duke University suggests that utilities often operate below their potential capacity, indicating that energy efficiency improvements could alleviate some pressure without resorting to building new fossil fuel infrastructures. However, the prevailing narrative prioritizes immediate energy demands over long-term environmental sustainability. With substantial investments flowing into cleaner energy technologies, such as solar and nuclear, there is hope for a transition away from fossil fuels. However, for now, the communities hosting these new data centers face significant challenges, both financially and environmentally, as they grapple with the consequences of developments they never chose to welcome.