Most Americans Say AI Is Bad for Society, Oppose More Data Centers

The thread mostly rejected the idea that widespread AI use proves public approval. Several commenters argued that adoption is a poor proxy for consent when AI features are embedded into search, television interfaces, workplace tools, and other services people need to function. They compared this to relying on environmentally damaging transport or joining an arms race: use can be compelled by circumstances even while people dislike the broader social outcome. One participant described stepping away from consumer AI as a relief, preferring ordinary web search and human-authored material to formulaic, flattering LLM output. The strongest concerns were concrete rather than abstract. Commenters tied hostility toward AI data centers to local electricity costs, water consumption, fossil-fuel backup generation, noise, diesel pollution, and weak local economic returns relative to their footprint. They also argued that AI leaders have helped create the backlash by foregrounding job displacement and catastrophic-risk narratives while appearing to concentrate gains among wealthy technology firms. A smaller group pushed back by treating rapid AI adoption and the public’s reliance on data-center-backed services as evidence that the technology delivers real value. Others said the debate is saturated with misinformation and political manipulation, but that dismissal drew criticism as an excuse for ignoring legitimate objections. Overall, the discussion leaned sharply skeptical of AI’s social governance and infrastructure costs, not necessarily of every useful AI application.