The gap between how AI executives see their industry and how the public sees it keeps widening, and the diagnoses coming out of Silicon Valley suggest the leaders building the technology still misunderstand why. A Pew Research Center survey published this week found more than 50% of Americans under 30 say they are more concerned than excited about AI, a figure that has grown 24% over the last five years. More than 70% of respondents across all age groups told Pew they believe AI will lead to fewer jobs.
Those numbers frame a problem that Meta CEO Mark Zuckerberg and Anthropic CEO Dario Amodei have each tried to answer publicly in the last two weeks, arriving at conflicting conclusions. Zuckerberg published a 6,500-word essay earlier this month laying out an optimistic vision in which every person has a personalized AI assistant that knows them intimately and helps with careers, hobbies, and childrearing. The product he describes does not yet exist.
Zuckerberg's essay leaned on the argument that giving everyone unfettered access to highly capable AI is the best-case scenario, and that better products will shift the mood. That framing treats the problem as one of capability and access rather than trust — a premise that the polling does not support. The Searchlight Institute found that hearing different messages about AI made little difference in how people rate the technology's societal impact.
Key facts
- 01More than 50% of Americans under 30 say they are more concerned than excited about AI, per Pew Research Center published this week.
- 02Concern over excitement has grown 24% across the last five years in Pew's tracking.
- 03More than 70% of respondents across all age groups told Pew they believe AI will lead to fewer jobs.
- 04Mark Zuckerberg published a 6,500-word essay this month arguing everyone should have a personalized AI assistant.
- 05The Searchlight Institute found that varying industry messaging made little difference in public opinion on AI's societal impact.
Zuckerberg also took a not-so-subtle shot at Amodei in the same piece, writing that he was surprised that discourse from many developing AI is so filled with doom. Tech investor Gavin Baker, appearing on the All-In podcast, went further, saying Dario's messaging has been massively helpful to efforts to ban data centers in America. White House AI adviser and All-In co-host David Sacks piled on, arguing that Anthropic's narratives have done more than anything to shape public fear.
Amodei answered with a lengthy social media post rejecting the idea that Anthropic's safety messaging is the cause of the backlash. He located the problem in a deeper place — a collapse of public confidence in institutions that predates the AI boom by decades.
Amodei wrote that ordinary people don't trust companies, governments, or the tech industry and always suspect the industry is cooking up some new way to screw them over. That reading matches the polling better than the messaging-problem thesis. Beyond jobs and data centers, Americans also tell pollsters they worry AI will harm young people's ability to form meaningful relationships, that tech companies should be required to compensate creators whose works were used to train models, and that AI could damage students' critical thinking skills.
“I think it is fundamentally a crisis of trust”— Dario Amodei, Anthropic CEO
Where Amodei loses the thread is on what to do about it. He argued in the same post that the most accurate criticism of AI companies is that they have not yet delivered on their big promises to benefit the world, citing the still-unfulfilled goal of using AI to cure cancer.
That framing implies the trust gap closes once the miracles arrive. But the polling suggests something different — that the backlash is being driven by what AI is already doing to society, not by what it has failed to do. Data centers going up in local communities, creative work being ingested without permission, and jobs already being reorganized are present-tense grievances, not future disappointments.
“the most accurate criticism of AI companies including Anthropic is that we haven't yet delivered on our big promises to benefit the world”— Dario Amodei, Anthropic CEO
The commercial landscape makes the point sharper than the polling does. Beverage brands Liquid Death and Garage Beer teamed with former Philadelphia Eagles player Jason Kelce on a spot inviting people to pee in bottles and send them to AI data centers. Liquid Death's creative lead told Marketing Dive that the data center backlash was literally the one thing that unites all Americans right now. When a beer ad reads the room better than the CEOs building the product, the industry has a positioning problem that cannot be solved with another essay.
Tech companies have spent heavily to make physical data centers more digestible, with commitments around water usage, clean energy, and local investment. It has not moved the numbers. The build-out has become the physical thing that people can point at when they want to express a broader unease about how tech companies engage with the communities they operate in.
For the AI market, the practical stakes of this misdiagnosis show up in adoption curves and regulatory exposure. If Zuckerberg is right that better products will win the argument, Meta and its peers can keep shipping and let the technology do the persuading. If Amodei is closer to right — and the polling suggests he is — then no product launch fixes this on its own, and the companies that treat trust as an engineering problem rather than a governance one will find their consumer businesses capped well below the total addressable market their investor decks assume.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
