Recently, a well-respected Catholic publication ran the first in what it billed as a series of three articles. Having recently returned from the Association of Catholic Publishers membership meeting, which featured three breakout sessions on artificial intelligence, one of my coworkers read the article and thought that something seemed a bit off. He thought he detected elements of AI writing, and taking a quick look, I agreed. I copied the article and pasted it into an AI tool designed to detect AI-generated content. The tool determined that 82% of the text was AI-generated. The only “original” material was an indented quotation from an encyclical by Pope St. John Paul II and a single sentence introducing that quotation.
Nowhere in the article or credit line was it noted that AI was used to generate the content. There was no editor’s note explaining its genesis, which leads me to believe that the editors of the publication had no idea what the person whose name was attached to the article had submitted. And the putative author himself had provided a lengthy bio in which he noted that the topic of the three-article series was near and dear to his heart.
The importance of a Catholic education
What was the topic? The uniqueness of a Catholic education.
The irony, dear reader, will likely not be lost on you. I don’t know which large language model the putative author had used to generate the text into which he had inserted a single quotation and one line of his own writing, but I do know this: There’s not a single LLM that has actually attended a Catholic school or benefited from a Catholic education. Trained on massive quantities of text — much of which was used, as anyone who has followed the development of these LLMs knows, without the knowledge or consent of the authors — the AI chatbot churned out a passable though stilted six-paragraph essay on the uniqueness of something it had not experienced. And the putative author, apparently without changing a word, presented this text as his own.
I, too, believe in the uniqueness of a Catholic education, even though I had no experience of Catholic education until my graduate studies at The Catholic University of America. Am I, perhaps, being hypocritical in criticizing an AI chatbot for “writing” about something it could never have experienced when my own experience has been so limited? The difference, of course, is that I can take my limited experience and imaginatively extrapolate from it to come to understand the benefits of a Catholic education; and I can look at the difference a Catholic education has made in the lives of people I know and admire. I can compare that with my own non-Catholic primary and secondary education and recognize what I missed.
I grew up in a small town in the middle of the Midwest in the ’70s and ’80s, and the conservative moral culture of our community was mirrored in our public schools, so my public education was, by today’s standards, a moral one. Yet, that morality was not situated in the context of the Catholic faith, and that is something that I could understand I missed, while an AI chatbot could not. Of course, the putative author of this LLM-generated essay presumably learned during the course of his Catholic education that passing off text that he did not write as his own was a violation of the spirit of the seventh commandment, even if legally it is impossible to steal from a chatbot, since AI-generated content is not copyrightable.
In other words, the mere fact of having experienced a Catholic education does not guarantee that one will conform one’s life to the lessons he has learned therein.
To say that the explosion of generative AI has launched us into a brave new world has become a cliché, but clichés are clichés because they are self-evidently true. There’s great potential in AI for making technology more useful, but great danger as well if we use it in ways that erode the moral and creative faculties of the human soul — those faculties that most fully reveal that we were created in the image and likeness of God.
