The gap between Silicon Valley’s AI ambitions and public sentiment is widening into a chasm. Pew Research dropped a number this week that should rattle every AI executive: 52% of Americans say they’re more concerned than excited about AI in daily life, a 15-point jump from 37% in 2021. That’s not a soft signal. It’s a trend line pointing the wrong direction for an industry that has raised hundreds of billions on the promise of inevitability.
The discontent is now bleeding into politics and business. The National Republican Senatorial Committee sent a memo to top AI companies warning that data center construction is hurting the party’s chances in an Ohio election. Meanwhile, The Wall Street Journal reported that tech companies are quietly sweetening data center deals to buy local goodwill — job guarantees, clean water investments, and in one Louisiana parish, $50,000 bonuses for teachers. When you’re paying six figures to calm down a school board, you have a perception problem.
Airbnb CEO Brian Chesky put it bluntly on a recent podcast: the industry isn’t shipping products that “regular people” actually like. Anthropic CEO Dario Amodei went further, calling the negative public perception a “big problem” and a “crisis of trust” on X. He acknowledged what most executives won’t say out loud — people “suspect that we are cooking up some new way to screw them over.” His proposed fix? Deliver on the big promises, like curing cancer. That’s a tall order for a quarterly earnings cycle.
What’s striking is the consumer behavior underneath the polling. Young people are buying dumbphones, point-and-shoot cameras, and AI-free iPods at premium prices on eBay. Run clubs are replacing dating apps. Quilting and Mahjong are having a moment. This isn’t a messaging failure — it’s a value proposition failure. Consumers understand AI well enough to know they’re being asked to absorb the costs (job anxiety, IP theft concerns, chatbot-saturated products) without seeing the upside. The industry assumed adoption would breed acceptance. The data suggests the opposite is happening.
