Select Page

EPISODE 22  ·  October 28, 2026  ·  On Message: AI Pushback

On Message: AI has a message problem

AI's future depends on more than better technology. This On Message analysis examines data-center concerns, the limits of sweeping predictions, and how Utah's pro-human AI framework connects potential benefits with public responsibility.

About this Episode:

The AI race is also a communications race. Building extraordinary technology won’t automatically persuade people to welcome it, especially when the promises feel distant and the disruptions feel close.

That’s the central argument in this On Message analysis: AI leaders need to explain what gets better, how people are protected, and who shares in the gains. For Utah’s policymakers and business leaders, the question isn’t simply whether to embrace AI. It’s how to earn the confidence needed to put it to work.

Public trust starts with questions, not predictions

For a community considering a data center, AI isn’t an abstract debate about the future. It’s a proposed development with questions about electricity, water, noise, and who pays. Those questions deserve specific answers, not a lecture about why technological progress is inevitable.

Blocking one project won’t stop AI. But that doesn’t erase a community’s right to question the project. Winning a permit and winning public trust are different accomplishments.

The same problem runs through the industry’s broader messaging. Catastrophe and utopia make attention-grabbing headlines, but neither tells a parent, patient, or business owner what changes today. Leading with benefits doesn’t mean concealing risks. It means explaining the opportunity and the protections together.

The essential follow-up is: How?

Sweeping claims about AI need the same scrutiny whether they promise abundance or predict disaster. Which tasks could disappear? Which workers are affected? Over what period? If healthcare improves, what gets better and who remains accountable?

Those aren’t anti-technology questions. They’re the questions that turn a prediction into something people can evaluate. A frightening forecast isn’t proof, and a dazzling promise isn’t a plan.

Consider the promise of giving people time back. Less administrative work could help a small-business owner serve customers or help a small team compete. But efficiency doesn’t automatically produce shorter workdays. Someone must decide how the gains are shared. The human benefit isn’t more time using AI. It’s more time for life.

Utah offers a people-first starting point

Governor Spencer Cox’s pro-human AI framework offers a useful way to organize that conversation. Its value as a communications model is the connection between potential benefits and responsibility, rather than allegiance to a technology for its own sake.

In education, the framework emphasizes human agency and mentorship. The message is straightforward: help teachers teach and students learn. In healthcare, it pairs opportunity with transparency, accountability, safety, and patient protection. For workers, it points toward apprenticeships and hands-on re-skilling.

Those commitments don’t establish that every initiative will succeed or that public opinion has improved. They do provide a more concrete starting point than paradise versus extinction: What gets better? What protects people? How can they participate?

Compete on products. Cooperate on trust.

AI companies can compete fiercely for customers while recognizing a shared interest in public confidence. The industry needs a playbook built around five commitments: lead with people, explain how, earn trust, share the gains, and stay on message.

That means replacing sweeping predictions with evidence and answering questions about costs, resources, privacy, and safety. It also means investing in skills and visible community benefits. Philanthropy can reinforce responsible operations, but it cannot substitute for them.

Message discipline matters only if the underlying conduct supports the message. Repeating a promise won’t fix an unanswered question or an unfair arrangement. Leaders need to test what people understand, address what concerns them, and change course when the concern is justified.

Capital can build computing power. It cannot purchase public confidence by itself. AI’s pioneers can build the technology. Can they earn the public’s trust?

Full Transcript

The AI race is also a communications race. Building extraordinary technology won’t automatically persuade people to welcome it, especially when the promises feel distant and the disruptions feel close.

That’s the central argument in this On Message analysis: AI leaders need to explain what gets better, how people are protected, and who shares in the gains. For Utah’s policymakers and business leaders, the question isn’t simply whether to embrace AI. It’s how to earn the confidence needed to put it to work.

Public trust starts with questions, not predictions

For a community considering a data center, AI isn’t an abstract debate about the future. It’s a proposed development with questions about electricity, water, noise, and who pays. Those questions deserve specific answers, not a lecture about why technological progress is inevitable.

Blocking one project won’t stop AI. But that doesn’t erase a community’s right to question the project. Winning a permit and winning public trust are different accomplishments.

The same problem runs through the industry’s broader messaging. Catastrophe and utopia make attention-grabbing headlines, but neither tells a parent, patient, or business owner what changes today. Leading with benefits doesn’t mean concealing risks. It means explaining the opportunity and the protections together.

The essential follow-up is: How?

Sweeping claims about AI need the same scrutiny whether they promise abundance or predict disaster. Which tasks could disappear? Which workers are affected? Over what period? If healthcare improves, what gets better and who remains accountable?

Those aren’t anti-technology questions. They’re the questions that turn a prediction into something people can evaluate. A frightening forecast isn’t proof, and a dazzling promise isn’t a plan.

Consider the promise of giving people time back. Less administrative work could help a small-business owner serve customers or help a small team compete. But efficiency doesn’t automatically produce shorter workdays. Someone must decide how the gains are shared. The human benefit isn’t more time using AI. It’s more time for life.

Utah offers a people-first starting point

Governor Spencer Cox’s pro-human AI framework offers a useful way to organize that conversation. Its value as a communications model is the connection between potential benefits and responsibility, rather than allegiance to a technology for its own sake.

In education, the framework emphasizes human agency and mentorship. The message is straightforward: help teachers teach and students learn. In healthcare, it pairs opportunity with transparency, accountability, safety, and patient protection. For workers, it points toward apprenticeships and hands-on re-skilling.

Those commitments don’t establish that every initiative will succeed or that public opinion has improved. They do provide a more concrete starting point than paradise versus extinction: What gets better? What protects people? How can they participate?

Compete on products. Cooperate on trust.

AI companies can compete fiercely for customers while recognizing a shared interest in public confidence. The industry needs a playbook built around five commitments: lead with people, explain how, earn trust, share the gains, and stay on message.

That means replacing sweeping predictions with evidence and answering questions about costs, resources, privacy, and safety. It also means investing in skills and visible community benefits. Philanthropy can reinforce responsible operations, but it cannot substitute for them.

Message discipline matters only if the underlying conduct supports the message. Repeating a promise won’t fix an unanswered question or an unfair arrangement. Leaders need to test what people understand, address what concerns them, and change course when the concern is justified.

Capital can build computing power. It cannot purchase public confidence by itself. AI’s pioneers can build the technology. Can they earn the public’s trust?