I read Oren Cass’s New York Times article with great interest because he foregrounds something I have been sharing privately but deserves to be heard publicly.

Simply put, AI companies are not worthy of our trust pre-emptively. Or put another way: the onus rests on these companies to prove they are worthy of trust before anyone grants them trust.

Why? Cass lays out the case well: a fractured social environment, fragmented and degraded attention, and increasing isolation.

But the poet and philosopher Wendell Berry put it another way in his essay “Two Economies.” He notes that the prevailing market logic incentivizes “maximum profit or power with minimum responsibility.”

Is it any wonder that people are recoiling, not just against data centers, but a market logic that claims profit without responsibility?

When I have been talking with friends and colleagues, I am quick to distinguish between the potential value of AI “in theory” vs. “in the hands of max power/profit with min responsibility incentivized companies.” The last 20 years of Big Tech have borne out this skepticism.

The world’s most powerful people built products designed to capture our attention. Then they embedded these products into the infrastructure of everyday life. They became extraordinarily wealthy in the process. The costs of these products were also very real. Our attention is now something to capture and monetize. Moving social life off of platforms became increasingly difficult in proportion to the time and relationships moved onto them. Children and youth grew up profoundly (mis)shaped by these systems while all the while, executives feigned responsibility. Communities inherited misinformation, outrage, and isolation.

As a culture we trusted the people and platforms to be public goods and publicly good — connecting us across long distances, increasing human contact and relationships, democratizing access to knowledge and information. But the costs of the products never appeared in their accounting.

The irony of these unaccounted-for costs is that because the little economy sits in the Great Economy, “the loser’s losses finally afflict the winner.”

You can profit from fractured attention and still have to live in a culture with fractured attention. You can build the information environment and still have to raise your children inside it. You can push costs onto families, schools, communities, and civic life, but you do not get to leave the society those costs helped create.

The winner doesn’t get to live in a separate civilization.

So is it any wonder that people are recoiling not only from data centers or particular AI products, but from a market logic that continually claims profit and power while resisting responsibility for the consequences?

This is what Cass is seeing. We have spent the last 20 years watching the logic of the little economy work itself out before us: profit and power concentrated upward, while the costs were dispersed downward across our families, communities, and civic life.

This history ought to change the terms on which trust is given.

The reasonable posture is not to trust first and discover the costs later. It is to withhold trust until these institutions demonstrate that they are willing to accept responsibility for what they build, including the consequences that do not appear on their own ledgers.

We are a people who possess memory. Distrust, in this context, is not irrational resistance to technological progress. It is what memory should produce.

Trust should come only after trustworthiness.