A correct Catholic approach to AI becomes clearer, I think, if we approach a foundational text in Catholic Social Teaching, Rerum novarum, not as about structural issues in political economy, but rather as about claims on time and claims of authority.
The workhouses of the Industrial Revolution, by paying only a subsistence wage to the father, forced wives and children into the factories too, destroying time for the family, the parish, and worship. And making each member of a household directly dependent upon the owner, not the father. This configuration, moreover, appeared fixed; the members of a household seemed to have no way to escape their plight as “wage slaves.”
A “living wage” busts this up. Pay the father enough so that he can support a family and so that they, if they live thriftily, can acquire capital over time, and the result is that the family is restored as the basic cell of society. And the father’s authority too is restored.
Workhouses absorbed nearly all leisure time and took authority away from parents and clerics. The living wage, when honored, returned remunerative work to its proper position of being for the sake of the family, not the family for the sake of work.
Catholics face a situation today similar to the Industrial Age through what Harvard professor Shoshana Zuboff has called “surveillance capitalism.” Technology in the heady days of Wunderkinder like the young Steve Jobs and Bill Gates exulted to be in the service of the value creator – the entrepreneur, the artist, the executive looking for efficiencies of scale. But in roughly the early 2000s, things got flipped, so that the user became the product.
You know the maxim, “if the app is free, you are the product.” We pay for ostensibly “free” services not through money, but with our time and attention. If revenue comes from targeted advertising, then, once a network of users has ceased growing organically, further growth can come only from more screen time, or from more data, leading to better prediction and more assured control of behavior.
Furthermore, things get locked in. Get devices into the hands of children, and their behavior can be shaped into adulthood.
You see that your child is addicted to a screen? My colleagues around the country say that students can’t sit through a lecture any longer: they must “go to the bathroom” at least once an hour, a euphemism for going away to look at their phones – the way cigarette addicts used to behave. These failings are not accidents or mere weaknesses of human nature.
Are our clerics paying attention here? Christians are supposed to live “in the presence of God,” not in the presence of short-form videos. If we have free time, saying a prayer is a good thing to do, or visiting a church. Families are supposed to center around the fellowship among the children, not Instagram networks, and follow the culture set by parents, not influencers.
Priests and bishops who are internet celebrities are like worker priests who penetrated factories after the Industrial Revolution. They do good work, to be sure, but they are not naming the fundamental problem, or contributing to the needed change in our thinking about how technology uses us.
In particular, they are not helping to foment this other “paradigm change,” which Zuboff has rightly said is necessary to overcome “surveillance capitalism” – the way we came to see, as a society, that cigarette addictions and polluting the environment are to be shunned.
The main ethical question concerning AI chatbots, then, is not new. Will these new technologies serve as de facto fiduciaries, putting the genuine interests of the user first, or will they join forces with existing “surveillance capitalism,” so that chats come to be in the service of an advertising master besides the user; and users are drawn more deeply into a web of subjective illusion?
Only Anthropic among the leading companies has foresworn advertisement as a source of revenue. Anthropic also offers users clear options for excluding their data from model training, such as by “incognito mode.” Nothing, however, keeps Anthropic from changing its policy if, say, it faces financial stress in the future. Arguably, all AI chatbots – by regulation – should be required to follow Anthropic’s current business model.
Just as Leo XIII, on principles of natural right, ruled out communism as a solution to worker exploitation, so today Catholics must rule out, as a solution to “surveillance capitalism,” the communism of social life evident in China. In China, the state moderates children’s screen time, not parents, and at the price of a system of social identification and control that is incompatible with religious, economic, and political liberty in a free society.
Every skilled person I know who uses AI claims a dramatic increase in their productivity, using AI not for entertainment but in the manner of a fiduciary. Apparently, Google has seen an across-the-board 10 percent increase in the efficiency of its engineers. For professors like me, the large-language model Claude can have the value of a part-time teaching assistant and research assistant together.
A simple rule, then, for a young person’s benefitting from the disruption which is certain to come would be: “Become through your education someone who is positioned to make good use of AI rather than be replaced by it.” The rule implies becoming, as much as possible, continually creative, a “source of content,” independent in thought, and having deep personal resources. The profile of a sociable entrepreneur would do the trick – or the youngest child in a large family.
The attainment of such intermediate goals requires, for most of us, a serious return to the wellsprings of creativity in Western civilization, in fellowship with like-minded others.
That is not a bad thing at all. It will require families that are “domestic churches,” parishes that care about mystery and doctrine, real habits of prayer among the Lord’s disciples, and schools that are genuine communities of seekers after the transcendentals of truth, beauty, goodness, and unity.
