Old Texts, New Tech
Notes on Postman, the Pope, and AI
…To the young, schooling seems relentless, but we know it is not. What is relentless is our education, which, for good or ill, gives us no rest. That is why poverty is a great educator. Having no boundaries and refusing to be ignored, it mostly teaches hopelessness. But not always. Politics is also a great educator. Mostly it teaches, I am afraid, cynicism. But not always. Television is a great educator as well. Mostly it teaches consumerism. But not always.
It is the “not always” that keeps the romantic spirit alive in those who write about schooling. The faith is that despite some of the more debilitating teachings of culture itself, something can be done in school that will alter the lenses through which one sees the world; which is to say, that non-trivial schooling can provide a point of view from which what is can be seen clearly, what was as a living present, and what will be as filled with possibility.
What this means is that at its best, schooling can be about how to make a life, which is quite different from how to make a living. Such an enterprise is not easy to pursue, since our politicians rarely speak of it, our technology is indifferent to it, and our commerce despises it. Nonetheless, it is the weightiest and most important thing to write about.
—from the preface to Neil Postman’s The End of Education (1996)
In my last post of “Teacher/Poet” I introduced media theorist Neil Postman’s “10 Principles for Technology Education” and set about close reading one of them, “#3,” and then offered the reader the full list at the end of the post, promising to write a sequel soon, again applying these principles to generative AI, and especially considering generative AI’s use (or non-use) in schools. Here, reader, is my fulfillment to that promise, part two.
That meaty quote from the preface of The End of Education, the book where the “10 Principles” were first published, gives you a helpful frame for interpreting where Postman is coming from in his delivering us these principles. “At its best,” he writes in 1996, “schooling can be about how to make a life, which is quite different from how to make a living.” Indeed. But in 2026, how relentless the push is, and from so many people in power, to incorporate AI into the schooling of all of our children so that they are “ready for a changing workplace.” Ready, not to make a life, but to make a living.
I’ll argue at the end of this essay that this utilitarian strain has actually always been present (and often dominant) in American education—and I don’t think Postman pays enough attention to it. At the same time I cannot help be inspired by his consideration of schooling and learning. As the old biblical Proverb has it, “Where there is no vision, the people perish.” (29:18.) We’ve got plenty of technicians working the vineyard, very few visionaries. I’m delighted to deepen your introduction to one such visionary here, reader. Let’s get back to that close reading of his “10 Principles,” starting from the top.
1. All technological change is a Faustian bargain. For every advantage a new technology offers, there is a corresponding disadvantage.
2. The advantages and disadvantages of new technologies are never distributed evenly among the population. This means that every new technology benefits some and harms others.
—Neil Postman’s First and Second Principles of Technology Education
In the Faust play that I have read, Marlowe’s, I don’t remember the title character thinking so much about humanity. Or if he did, only insofar as those humans affected him, and his concerns. This tracks with the creators of generative AI and LLMs, in spite of tech bros’ claims to the contrary. The biggest beneficiaries of the “AI Boom” right now are the companies who have created the technology. Other temporary beneficiaries may include people with expertise and education in a particular field—the more advanced, the better. Some of these report learning to do more and faster with that knowledge and expertise when aided by these new products. But the less educated a user of an LLM is, the more likely an LLM seems to lead the user astray or even hinder cognitive development. (And for younger folks, learning to write, recent studies show that excessive LLM use can be devastating.)
As there is clearly a correlation between income and education, AI seems to be harming poorer and marginalized communities at a much higher percentage than it is affluent ones. In addition, the environmentally and community-damaging data centers needed to power the “boom” are being built in economically depressed neighborhoods. AI as a tech phenomenon is exacerbating existing inequalities: The rich are getting richer; the poor are getting poorer.
3. Embedded in every technology there is a powerful idea, sometimes two or three powerful ideas. Like language itself, a technology predisposes us to favor and value certain perspectives and accomplishments and to subordinate others. Every technology has a philosophy, which is given expression in how the technology makes people use their minds, in what it makes us do with our bodies, in how it codifies the world, in which of our senses it amplifies, in which of our emotional and intellectual tendencies it disregards.
—Neil Postman’s Third Principle of Technology Education
I wrote on this principle in my first post, so those interested in a deeper dive can check that out, but these two paragraphs offer a distillation of what I think he’s up to:
“To stick with Postman’s language, what ‘perspectives and accomplishment’ are LLMs ‘predisposed to favor and value’? And which are they predisposed to ‘subordinate’? Efficiency comes to my mind right away as an accomplishment that LLMs value highly. Tasks that previously took large amounts of time can now be done in seconds. Productivity goes hand in hand with this value: Because LLMs can do so much and so quickly, they allow users and companies the capacity to produce more, and with less human labor. What is subordinated, then, in this embrace of LLMs is the distinctive value of human labor, which is often messy, has fits and starts, and sometimes goes nowhere at all before it arrives at its destination.
The product—the thing you want accomplished when you set out to start your labor—is valued immensely by the LLM and by LLM users. Process—how a human being labors to create or make the product is devalued steeply, so steeply in fact that many human beings are already losing their jobs in the face of this technological revolution.”
4. A new technology usually makes war against an old technology.
—Neil Postman’s Fourth Principle of Technology Education
There may very well be other “old” technologies AI is “warring against,” but I am most interested in this essay in exploring how AI is at war with books.
At war against a book as a physical, time-bound thing. As a representation of unique and singular voices (though of course not all books or perhaps even any are truly “singularly” authored). At war against this artifact that is a record of its own making, and includes the many people and communities that went into its making.
AI companies like Anthropic and OpenAI took millions of these artifacts and forced the LLMs they have created to “ingest” them. Post-ingestion, though the LLM might benefit from processing the particular voice, style, and “content” of the book, all that went into the making of the book has been erased. Flattened.
From an intellectual property/copyright standpoint, because the companies purchased physical copies of the books that the machines later “ingested,” it’s possible the companies could get away with what is clearly theft. But regardless of how these courtroom struggles play out between the authors and publishers and the AI companies, the Postman point holds: These are battles in a larger war that the new technology is waging against an older, more textured one. In Postman’s language, AI competes with the book for “time, attention, prestige, and ‘worldview.’”
The reader of this blog might well say that books were already being displaced by newer technologies, like the Internet. Surely this was true, but to my mind the current—and unapologetic—thefts of intellectual property mark a new kind of aggression. The ingestion of authors’ works into LLMs deprives these authors of both money and prestige. And because AI is being pushed by so many in power into so many aspects of our lives, it is also front and center in the time and attention we must give and pay it. With so much time and attention lavished on AI (no, the new tech has not “freed us up” to do more, as promised), there is less time and attention for that older and more leisurely act, the act of reading.
5. Technological change is not additive; it is ecological. A new technology does not merely add something; it changes everything.
—Neil Postman’s Fifth Principle of Technology Education
Reading over what I’ve written on principle #4, especially in light of conversations I’ve heard between historians Matt Seybold and Jeff Jarvis around the introduction of print with Gutenberg, I’m inclined to complicate some of what I’ve made seem like a “book good / AI bad” binary, which I admit I can lean into, and even moreso now that I am actively publishing books, and for profit.
In one of these conversations on his excellent podcast, American Vandal, Seybold pointed out, in an offhand manner that the act of publishing a book is itself a privatizing and a commodification of ideas and imagination—in other words, something I criticize about AI, just an older form of it. And as Jarvis said in that conversation, (“The Gutenberg Parenthesis” was the episode, I believe) the truly revolutionary technology of our era, which is still relatively young, is the internet, a technology that has had massively democratizing effects (though, since the AI boom, as Seybold and other guests to his program have noted, has been significantly degraded, or “enshittified,” to again use Cory Doctorow’s term.)
LLMs really wouldn’t be the powerful and impressive machines they are without the massive amounts of data they have access to from the internet. (See Audrey Watters’ brilliant Teaching Machines for accounts of much less impressive but no less dogged attempts by technologists like B.F. Skinner to embed forms of “artificial intelligence” into K-12 classrooms and universities in the middle of the 20th century. Teachers and schools then were much more inclined to resist, where now they are folding to Silicon Valley.) The internet really did, to use Postman’s phrase, “change everything,” and Jarvis argues we are still catching up to that change.
Was and is the internet as a technology “at war” with books? Perhaps not as openly as AI is as a technology, but I do think it is and was. As I started writing this paragraph, I was tempted to say earlier versions of the internet were like these beautifully gargantuan world-spanning libraries. Except if that metaphor were to hold then every time you touched a book, let alone checked one out, the library would take note, mark your habits and interests, monitor them, store them in a cloud. Except if the metaphor were to hold, the library wouldn’t be free but an indispensable engine of capital, always trying to sell you on something, hook you on some new product….No the metaphor doesn’t hold.
6. Because of the symbolic forms in which information is encoded, different technologies have different intellectual and emotional biases.
—Neil Postman’s Sixth Principle of Technology Education
“Symbolic forms” is an important phrase here. In relationship to what I’ve been mulling over in this essay, it seems worth asking the question: what are the “symbolic forms” that information is encoded in in books versus the internet versus an LLM or chatbot? I’ll give it a shot. The book: words on a page, between covers. The book: can be held in one’s hands, given as a gift. Thrown against a wall in frustration. Ignored for years. The book: can gather dust. The internet: a website with links. The internet: multiple ways in with the eyes—video and text and photographs. The mouse and your finger working together to scroll. The touchscreen on your phone. The internet: portal to your wallet. The internet: jumbled space—so much of your logistical and imaginative and intellectual and capitalistic life rolled into one. You buy food and clothing and watch movies and catch up with friends and everything in between. But the “symbolic forms”? These massive companies—Facebook/Meta, Twitter/X, Google, etc.—they shape all these experiences, they gather the data on you. Surveillance has always been at the heart of the internet as a form. And with an LLM this extends. Chatbot: mirror to your desires? Spooler of text at command? Your personal sycophant-in-chief?
As I tease out what the “symbolic forms” of book versus internet versus LLM might be, I consider too the second part of Postman’s sentence: “Because of the symbolic forms in which information is encoded, different technologies have different intellectual and emotional biases.”
The bias of the book as a technology seems to be towards absorption, immersion. If you are reading a book and you resist absorption and immersion, you will probably put the book down, do something else. But the internet and the attendant technologies (social media, and now AI) that comprise it, encourage, rather than absorption, a kind of restlessness. Jumping here and there, with momentary frissons of emotion that lead to other momentary frissons of emotion. It’s hard to go on the internet (definitely on a social media platform) and simply “find the thing you’re looking for.” You don’t close up the browser with the same ease and finality that you close up a book. We finish books. We never finish living our online lives.
7. Because of the accessibility and speed of their information, different technologies have different political biases.
—Neil Postman’s Seventh Principle for Technology Education
It’s important to remind myself as I write these reflections that Neil Postman wrote these 10 principles in 1996, when the Internet was still a very young technology. At the time, at least as teenage me remembers it, there was a sense of this vast “openness” to the web. It seemed to be accessible to anyone, provided you had the means to connect to it. (“having the means to connect to it” is a pretty big caveat though, now that I’ve written that phrase out.) But in terms of “political bias,” the internet seemed, like books, to resemble a pretty deeply democratic technology. The passage of time has surely complicated such a simplistic understanding, and our current fascist American administration complicates that understanding even more. Trump and Co. have made great use of (and been used by) Silicon Valley titans who reveal an internet that is less and less a commons for public benefit and more and more a privatized engine for commerce flowing to particular companies—often the companies of the titans themselves.
When it comes to AI as a technology, the impacts on labor specifically are profoundly undemocratic. The emphasis of this technology on accessibility and speed has meant companies cutting the jobs of many people, and increasing the precarity and uncertainty for many more.
For my own particular labor—as a teacher, specifically as a teacher of reading and writing—I am struck by the ways in which educational technology’s weaving itself into the infrastructure of schools has muted teacher’s resistance to this technology. That is, since there are already many, many educational technologies that schools mandate teachers use on a daily basis, “folding in” this additional tech (generative AI) into other “Learning Management Systems,” as it has been folded in with Canvas, for instance, makes it harder for teachers to refuse the use of it, even when there are already studies suggesting prolonged use may do cognitive harm to young people. (Let me link you again to MIT’s June, 2025 study, “Your Brain on ChatGPT”)
8. Because of their physical form, different technologies have different sensory biases.
—Neil Postman’s Eighth Principle of Technology Education
I’m inclined to say the internet, the LLM, the chatbot, and the book are all in their way “biased” towards the visual. Though with each of these technologies there are important oral and auditory qualities, too. With the internet, there are so many videos benefiting equally from sound as sight, and these are of course aggregated by AI companies now as well. With books, we know there has been a boom of audio books recently and humans have often delighted in the act of reading things aloud. Still, the visual does seem like the primary sense here. So, if we take this principle along with #4, there does seem to be a real sense in which the internet and LLMs and chatbots are “warring” with books for our eyeballs’ attention.
I am no literacy expert. However, I’ve read and experienced enough in twenty years of teaching high school English and living a literary life to know that the kind of reading you do online or on your phone is often very different in quality than the kind you do with a book in your hand. As most people understand intuitively, the online kind of reading is more likely to lend itself to utilitarian or extractive modes (get what you need for the thing you’re doing and move one) rather than the immersive ones I mentioned earlier. And the kind of processing that LLMs and other AI models do, is of course not reading at all.
If you consider my insight in #6, about the ways the internet and the LLM interact with our attention in contrast to books, then it’s probably true that in the “war for our eyeballs” these technologies propose radically different ways of “seeing.”
9. Because of the conditions in which we attend to them, different technologies have different social biases.
—Neil Postman’s Ninth Principle of Technology Education
As you’ve noted in other pieces here, I am unapologetically reading generative AI through the very specific lens of a high school English teacher. So, for me, the “conditions” to use Postman’s word, relate specifically to the institution of the school, particularly the high school. In my experience and estimation, as it relates to high school, AI as a technology is biased towards atomization and conformity. Atomization, because the technology leads to hyper-individualized responses to the users’ prompting, and so has the social effect of lessening interactions with actual human beings (other students and teachers), thus weakening community bonds. And conformity, because use of LLMs encourage a replication of what has been done before, rather than the trial or testing of something new.
In addition, the mandate (coming from the very top of the United States Government) to integrate AI into all K-12 schools exacerbates this conforming effect: Teachers and students “refusing to play” the AI game could find themselves ostracized from a class activity or assignment (in the case of a student) unless the scaffolding and communication from the teacher is handled extremely thoughtfully, or (in the case of a teacher) ostracized from fellow faculty and administrators or outright fired for refusing to use the technology. In my estimation this situation fits hand-in-glove with Postman’s Seventh Principle: AI is uniquely well-suited to a powerful fascist government, the one we are experiencing now in the United States in 2026.
Postman himself noted in his long career reflecting on education that there are times when teaching should be a “conserving” activity (see his 1979 book, Teaching as a Conserving Activity) and times when it should be “subversive” (see his 1969 book, Teaching as a Subversive Activity). I’m no “Postman-whisperer” and don’t claim in this essay to speak for him. I do think, though, for those teachers aware of the fascist turn in our politics and the way it has deformed our classrooms and schools, subversion is a more appropriate path than conservation at this particular moment. Or, perhaps to put it in a more nuanced way, it is a moment to both conserve and subvert: conserve time-tested practices for learning (including the reading of actual books), and subvert the ill-conceived corporatization of our institutions (via, in one way, wholesale “integration” and incorporation of generative AI into everything).
10. Because of their technical and economic structures, different technologies have different content biases.
—Neil Postman’s Tenth Principle for Technology Education
Karen Hao has pointed out in her excellent book, Empire of AI, that the technology of generative AI does not necessitate the kind of hyperscaling done by companies like OpenAI and Anthropic, nor the kind of “everything machines” like ChatGPT. In fact, in concluding Empire of AI, she offers the example of a Maori community indigenous to a region of New Zealand that is using generative AI—with the full and transparent consent of the community—to preserve a language that might otherwise have been lost. This kind of use and preservation is immensely valuable to that particular community and as an exemplar to all human beings. But while it is immensely valuable, it of course is not nearly as financially profitable. The “content bias,” again to use Postman’s terms, is against focused, limited-in-scale projects like this one and for “everything machines” like ChatGPT—a product that scrapes vast amounts of the internet and published human thought and writing (without permission) to feed its creators’ desire for profits and control.
Books, of course, aren’t “pure” either, when we consider the “content” between their covers. Many books have been and will continue to be written simply to sell copies. And most publishers (myself included) have a commercial interest in the books they put out into the world. We want the books to sell. But even the crassest, most commercialistic or sensationalistic writing is necessarily focused on a particular content area or topic, whether that’s celebrity gossip or how to make millions with sports gambling, or whatever. The “content bias” of print, then, while it isn’t necessarily tilted toward quality, is at least biased toward specificity, maybe even expertise. LLMs can certainly cannibalize this specificity and expertise (witness the already-mentioned massive theft of books by OpenAI and Anthropic) but the actual output of their machines is decidedly “mid” as Tressie McMillan Cottom has so memorably put it. The “voice “of the machine may attempt to imitate the writers it ingests but it doesn’t do anything new. While use of LLMs is provoking a scare for publishers like me, there is still the bedrock reality that actual readers don’t want to read writing produced by an LLM. They want human-produced writing, which they perceive as more valuable.
I noted in my last installment that as I have been revisiting Postman’s The End of Education, I have been listening to Matt Seybold’s wonderful podcast The American Vandal, which offers great insights into all of the “principles” touched on here, and in significant ways updates them to account for Edtech’s targeted incursions into schools and universities these past thirty years since Neil Postman’s book was published. Both Postman and Seybold share a profound distrust of the corporatization and financialization of so many of our civic institutions. Since I’m putting the two of these thinkers in conversation with each other in this series, I should note that Postman, at least in 1996, didn’t seem to see schools and universities as vulnerable to this process as Seybold does in 2026. Perhaps this is simply because these institutions in 2026 are much weaker after decades of being attacked? That may be, but I think it is also because Postman has a greater faith in what he calls the “American Experiment,” a myth (his word) about our national history which he finds helpful for schools, especially public schools, to propagate. This passage from The End of Education is representative:
“Public education does not serve a public. It creates a public. And in creating the right kind of public, the schools contribute toward strengthening the spiritual basis of the American Creed. That is how Jefferson understood it, how Horace Mann understood it, how John Dewey understood it, and in fact, there is no other way to understand it. The question is not, does or doesn’t public schooling create a public? The question is, what kind of public does it create?....The answer to this question has nothing to do with computers, with testing, with teacher accountability, and with the other details of managing schools. The right answer depends on two things, and two things alone: the existence of shared narratives and the capacity of such narratives to provide an inspired reason for schooling.”
Postman’s insight at the top of this section—that public education creates a public—is profound, as is his question about what might be the “right kind” of public. But (and I offer this “but” with some trepidation) I think the “American Creed,” as it relates to education, has always had some deep inequities baked in, and certainly ones that go back to Jefferson.
For the mass of Americans—and Jefferson wouldn’t speak against this—the purpose for education is that it is supposed to prepare citizens for the workforce. In modern parlance, that means it is a “pipeline” for the corporation, for business. This idea, which is anathema to folks like me (and Postman), is nonetheless present from the beginning with Jefferson, who though he may have venerated the ideal of a “citizen-scholar” for certain lucky and talented few, saw public education’s purpose as much more pragmatic for the rest of us. School, in America, was very often in practice a great sorting machine. This is the part of the “creed,” I suppose, that doesn’t get mentioned, or maybe is folded into a subordinate clause somewhere that you don’t pay much attention to. But it’s right there in Jefferson’s “Notes on the State of the Virginia,” first published in 1785, and quoted extensively in Harvard president James Bryant Conant’s 1940 essay, “Education for a Classless Society.” (I pull these dates to show the durability, and importance of Jefferson’s thinking. Conant in his commitment to standardization and standardized testing set the course for schooling that we continue to experience to this day.) Anyway here is the money quote from Jefferson’s “Notes on the State of Virginia,” where he turns his attention to schools and education:
“Another object of the revisal is, to diffuse knowledge more generally through the mass of the people. This bill proposes to lay off every county into small districts of five or six miles square, called hundreds, and in each of them to establish a school for teaching reading, writing, and arithmetic. The tutor to be supported by the hundred, and every person in it entitled to send their children three years gratis, and as much longer as they please, paying for it. These schools to be under a visitor, who is annually to chuse the boy, of best genius in the school, of those whose parents are too poor to give them further education, and to send him forward to one of the grammar schools, of which twenty are proposed to be erected in different parts of the country, for teaching Greek, Latin, geography, and the higher branches of numerical arithmetic. Of the boys thus sent in any one year, trial is to be made at the grammar schools one or two years, and the best genius of the whole selected, and continued six years, and the residue dismissed. By this means twenty of the best geniusses will be raked from the rubbish annually, and be instructed, at the public expence, so far as the grammar schools go.” [Emphasis mine]
The central creed—and thanks to Michael Sandel’s excellent book on the topic for showing this connection between Jefferson and Conant—is meritocracy. It is not that all Americans are entitled to an equal quality education. It is that the ones (note all boys and men) that are tested and tried as the “best geniuses of the whole” are selected and then moved onward to the next stage. If you demonstrate aptitude for great things, then, and only then, you move on to receive benefits that help you achieve for greatness. As for the rest of the students, as Harvard President Conant put it in his 1940 essay, “Education for the Classless Society,” referring to the “Jeffersonian Tradition,” testing was the central way to handle the “horde of heterogenous students” that “descended” on secondary schools throughout the United States. Harvard became the birthplace of modern standardized testing, and Jefferson’s vision for education was part of the rationale for why that happened. Jefferson and Bryant Conant can help us understand too why the corporatized (branded) College Board and its suite of meritocratic “Advanced Placement” tests have come to dominate and set the conversation for what high school education is and can be. It also helps explains why, in recent years, and even months, the company has pivoted explicitly to “workforce preparation” in its AP course offerings (AP Business with Personal Finance) and even in the way it presents all of its courses.
How does all of this connect to AI and schools, specifically in an American context, in 2026? I don’t think it is an accident that at the very same time Trump is pushing AI adoption in all public schools, kindergarten through and 12th grade, that there is a very popular and very privatized “Classical Schools” movement that offers young people a mostly tech-free education. This movement has even created its own rival standardized alternative to the College Board’s SAT—the “Classic Learning Test.” I also don’t think it’s an accident that some of the folks active in the shaping of this privatized alternative to public schools and schooling and who serve on boards for this movement, are actively looking to dismantle and hollow out education as a public good. For an exemplar of this, see Chris Rufo’s aptly titled address, “Laying Siege to the Institutions” made at Hillsdale College in April, 2022. One such institution he his cohorts have openly “laid siege to” is public education, from kindergarten through twenty-second grade. The gloves, and the masks, are off.
I will be the first to tell you that I am someone who benefited greatly from a “great books” education. And incidentally, my alma mater, University of Dallas, has recently made headlines as the first U.S. university to award credit for the Classic Learning Test and its offshoots. I see the value in this kind of learning: reading great books with good teachers, talking about them at the seminar table. It’s called “classic” for a reason—this transformational and important work and has proved valuable to human beings for centuries, maybe even millennia. And I do wish more public schools and universities would commit to reading “great books” (determine that “greatness” how you will with consultation of your community) and talking about them at the seminar table.
At the same time, as Chris Rufo’s influence eloquently testifies, it cannot be denied that the “Classical Schools” movement is flourishing at the same time, and in part because of Trump’s fascist government. The movement fits hand in glove with Trump’s1776 Project, with his “Make America Great Again” vision, and with his vision of privatizing everything under the sun—including education. His administration’s new school voucher program is very good for Classical Schools, and for any private school. It is also working to hollow out public education.
I myself teach at a private high school, and so my own institution would theoretically benefit from my state’s “opting in” (it hasn’t yet) to Trump’s voucher program. Speaking for myself, though, I oppose the use of public funds for private education. And as a teacher who has witnessed firsthand in 2026 how this fascist federal government utilized public funds to terrorize citizens of the city where I teach through a monthslong ICE occupation, I would encourage other private school teachers, especially ones who teach at Catholic schools, to stand in solidarity against these policies. After all, the same “Big Beautiful Bill” that earmarked all that money for vouchers also ramped up spending for ICE.
Solidarity can’t just be a sentence in a document, something that sounds nice but we would never consider acting on. I’ve been thinking about solidarity a lot this past week as I’ve been writing this piece, finishing up another school year as a teacher, and now reading Pope Leo’s new teaching document on AI, Magnifica Humanitas.
Throughout Magnifica Humanitas, Pope Leo makes it clear that we must view this technology through the lens of what he calls the “social doctrine” of the Catholic Church (he begins the encyclical with a pretty robust survey/summary of the last 130 years of Catholic Social Teaching from Pope Leo XIII all the way up to John Paul II, Benedict XVI and his immediate predecessor, Francis). Seeing through that lens of social doctrine for Leo means seeing this technology through the eyes of its victims.
Pope Leo would clearly co-sign Postman’s first two principles of technology education: “The advantages and disadvantages of new technologies are never distributed evenly among the population. This means that every new technology benefits some and harms others.” Pope Leo sees and names in this document that the current version of generative AI being deployed around the world has been built on the backs of the poor in exploitative ways (see paragraph 173), constitutes the threat of a “new colonialism” (see paragraph 178, is the Pope reading Karen Hao??), and is being weaponized to wage wars (see Chapter 5, 182-227). This is moving, and prophetic (in the biblical sense), and I would hope inspires not just Catholics but all people of good will to take action in big and small ways to “join the conversation” around how to use (and not use) this technology in responsible ways. And if the table where the conversations are happening is not being opened in an inclusive and democratic way, demanding the table, the conversation be re-shaped.
For Catholics like me there is also the awkward but I think necessary obligation to name as Pope Leo does in Magnifica Humanitas that the call towards this social responsibility, this solidarity, is not particularly new. Since Pope Leo XIII and Rerum Novarum in 1891, the Catholic Church has taught, for instance, that the dignity of the human worker must be respected, and that all workers must be free to form unions, to give a particular example that touch specifically on having a more “inclusive and democratic” conversations about AI. Because of course, there are many leaders—especially corporate ones—who don’t want these conversations. As Audrey Watters aptly puts it: “AI isn’t coming for your jobs. Your bosses are coming for your jobs.” If you are an individual, atomized worker sitting across the table from a manager who wants to automate everything with AI, you’re out of luck, and probably out of a job. If you’re collectively bargaining with a union supporting you, you have power.
I wrote a few paragraphs earlier “Solidarity can’t just be a sentence in a document, something that sounds nice but we would never consider acting on.” The sentences of Magnifica Humanitas about labor, unions, and workers have some bite, oblige us to do something. Pope Leo writes, “Labor unions, which the Church has consistently supported, are called upon to be open to new types of employment and the corresponding needs of workers, in order to represent and defend them.” (Par. 155). And then continues a few sentences later to say, “every introduction of automation and AI should be accompanied by verifiable measures to protect the employment, retraining and participation of workers. In this way, technology will be oriented toward freeing up human time and capabilities, rather than producing exclusion.” (Par. 156) Again, if these sayings by the leader of the worldwide Roman Catholic Church are to have any meaning, leaders of local Catholic institutions will have to actually put them into practice.
It’s all fine and good for Catholic leaders in the United States left, right, and center to laud Pope Leo’s message on AI and reiteration of support of the dignity of workers (as they have been doing these past few weeks). But questions do arise. For instance, why in the United States, a country with nearly 150,000 K-12 Catholic schoolteachers do only a few thousand (around 5% at most) enjoy the protections of a union? Why, when faculty members at Catholic universities organize in solidarity to support the most precarious of teachers—adjunct instructors—do leaders of those universities move swiftly to crush said organizing? (For a recent example of Catholic union-busting, read this 2024 National Catholic Reporter piece about Marquette University’s move to “squash unionizing efforts.”)
There are other ways. Three distinct unions for Rutgers University—representing as Academe Magazine put it, “tenured and tenure-track faculty, non-tenure-track instructors, librarians, adjuncts, graduate workers, postdoctoral students, researchers, and clinicians,” struck together in April, 2023. This meant actual solidarity—folks with more privilege and prestige standing in line with those more precarious. As a result, Rutgers’ adjuncts, while still enduring some of the same slings and arrows as others across the country, made historic gains: after collective bargaining, their minimum pay per credit hour is now $2,777 per credit hour, more than twice the national average of $1,166. (To be clear that increase is inadequate, but it is an achievement, step in the right direction.)
I mention the Catholic/Pope Leo connection in closing this essay because it highlights common ground that folks have with regard to AI that exists across typical left/right political divides in the US. We now know that resistance to building data centers to fuel the AI boom cuts across party lines. I would hazard a guess that “AI-fying” K-12 (or K-22) public schools as Trump has mandated also has broad resistance. The problem, of course, is solidarity. Many conservatives are now sending their children to private schools, a number of those (1500 strong and growing) are “classical schools.” For the classical schools, certainly, AI is not really an issue. You don’t do ChatGPT with Plato or Dante or Euclidean and Non-Euclidean Geometry. And you’ll have plenty of support if you just want to ignore Trump’s Executive Order nonsense on AI. A privilege your public school counterparts don’t enjoy.
As I was finishing this essay up last night, I just read a very thoughtful and incisive guest essay at Paul Weinhold’s “The Classical Ed Review” on the difference between “Artificial and Actual Intelligence” by director of the Director of Curriculum and Academic Resources for the Institute for Catholic Liberal Education (ICLE), Jake Tawney. One of Tawney’s insights here, that a “neural network” like artificial intelligence can’t “pursue truth, possess understanding,” or “apprehend essences” draws on Aquinas, but also on Neil Postman. Consider this helpful gem on how use of a technology shapes the user:
“Per McLuhan and Postman, the use of a technology will form us not merely by its misuse, but also by its mere use. Our interaction with A.I. will continue to shape the way that we think about intelligence itself, and the consequence may be that more and more people will come to define intelligence in terms of how a neural network functions.”
This is wisdom, for sure. For my part, I want to elevate and engage voices like Tawney’s. But also challenge them. In the political context we find ourselves in, where privatized alternatives are not simply proposed but done so with the intent to “lay siege” to public goods like education, it is not adequate only to make a case for “liberal education” in your own house. If your neighbor’s house and prospects are being laid waste, and you’re benefiting as a result, you have a responsibility to them as well. I’m not 100 % sure what is entailed by that responsibility in the long term, but in the short term it at least means meaningful dialogue with public school educators and leaders (and I think that would go vice versa for public school folks reading this), and consideration of what “movement growth” for your group might mean for the larger population as a whole.
NOTES:
-As I’ve mentioned, all of American Vandal is worth your time, but particularly for those interested in issues surrounding educational labor and AI, I recommend the episode “Theory at the Bargaining Table” with Anna Kornbluh and Dominique Baker. It’s focused primarily on issues that apply to higher education but certainly can be adapted for K-12 contexts.
-Another thinker whose work I’ve been drawing on these past months has been Astra Taylor, whose series of lectures, now in book form, The Age of Insecurity, makes a compelling case for solidarity and a public commons. And her chapter/lecture “Consumed by Curiosity” focuses thoughtfully and specifically on education. And her May, 2025 reflection on what a “universal right to education” might mean in the AI age is beautifully put.
This is a brilliant commentary, prophetic warning, and call to deep reflection -- and collective, purposeful action. I hope this finds the widest readership possible. Send it out; let it soar like the eagles. And let us pray that many others will see the wisdom you have shared.