Public opinion on artificial intelligence is increasingly turning negative, with communities protesting the construction of data centers and polls indicating that a majority of Americans believe AI is advancing too rapidly. This phenomenon, termed 'AI populism' by journalist Jasmine Sun, suggests that the resistance to AI stems not from concerns about its utility, but from anxieties about concentrated power and a lack of democratic input in its development and deployment.
Sun explains that AI populism is a worldview where AI is perceived not as a neutral technology, but as an "elite political project to be resisted." This perspective arises when individuals feel that large corporations and wealthy individuals are imposing changes on their communities without democratic consent. Unlike traditional AI safety concerns, which focus on technical risks, the populist backlash is driven by a "concentration-of-power, anti-elite dimension."
Evidence of this growing discontent has manifested in concerning ways, including attempts of political violence. The CEO of OpenAI, Sam Altman, was targeted by a Molotov cocktail and shooting incidents. In one instance, an Indiana councilman who voted for a data center project discovered gunshots at his home with a note reading "no data centers." Sun views these events as "warning shots" reflecting a broader trend of political violence and support for it, particularly among young people who feel they lack other channels for expressing their grievances. When institutional channels for recourse feel ineffective, opposition can become more diffuse and volatile, leading to actions ranging from online harassment to acts of violence.
Historically, technological shifts have been managed through established channels like labor unions. During the 20th century's automation wave, unions often negotiated terms for new machinery, ensuring workers benefited from increased productivity through wage hikes or shorter workweeks. In contrast, many industries impacted by AI currently lack strong union representation, and perceived democratic channels for addressing concerns are seen as inadequate. This historical parallel, reminiscent of the Luddite resistance to automated looms during the Industrial Revolution, highlights how the presence or absence of worker agency and social welfare policies influences the acceptance of automation. Economist Carl Benedikt Frey's research suggests that automation is more readily accepted when coupled with policies like higher minimum wages or redistribution.
Within Silicon Valley, there has been a noticeable shift in how seriously the AI backlash is taken over the past six months. Initially, reactions often dismissed discontent as temporary, similar to early reactions to the internet. However, growing opposition, including data center moratoriums and regulatory battles in Washington, has prompted AI leaders to acknowledge both messaging and product-related issues. Efforts by companies like OpenAI and Anthropic to propose economic policies addressing job automation reflect this heightened awareness. A significant attempt to preempt state-level AI regulation through federal legislation was also thwarted, demonstrating a bipartisan coalition's commitment to maintaining regulatory oversight.
In China, despite conditions that might foster similar populist sentiments—such as high youth unemployment and advanced AI adoption—a widespread backlash has not materialized. Sun attributes this to the Chinese government's paramount focus on domestic social stability, which suppresses dissent and limits a culture of resistance. Furthermore, for many Chinese citizens, technological advancement has historically been tied to dramatic economic progress and poverty reduction, fostering an outlook that emphasizes leveraging new technologies for economic gain rather than resisting them.
In the United States, a prevailing sense of pessimism about technology's impact may be fueling skepticism about AI's promises. This contrasts with the experience in China, where technological progress has often led to more concrete improvements in daily life. The perception of AI as a tool for individual advancement, rather than a threat, is more prevalent there. For example, in robot pharmacies and factories, efficiency gains and the ability to operate 24/7 are highlighted, addressing labor shortages and consumer needs, a contrast to the often-unspoken anxieties about job displacement in the U.S.
Looking ahead, AI populism is expected to play a more significant role in American politics. While its impact may be more visible in state and local races in the upcoming November elections, particularly concerning data center development, the 2028 presidential election is anticipated to feature AI as a central theme. Potential rises in unemployment linked to AI could amplify public concern. Additionally, politicians may leverage AI as a "convenient boogeyman" to advance their policy agendas, exploiting its current negative public perception and the perceived lack of public sympathy for AI billionaires to manufacture urgency for their proposals.