Avital Balwit, Chief of Staff for Antropic’s Dario Amodei

Avital Balwit’s essay on the Free Press titled “Searching for God in Silicon Valley” matters because it is not another outside critique of Silicon Valley. It is the testimony of an insider: the chief of staff to Anthropic’s Dario Amodei, someone embedded in the very world trying to build the future.

That alone gives the piece unusual weight. And her central insight is striking: the AI world is not simply secular, rational, and disenchanted. It has developed many of the working parts of a religion—conversion stories, eschatology, moral seriousness, schisms, and a sense of historical mission—while still failing to provide what religion gives a human being: consolation, humility, worship, and an answer to the question of how to live.

That is what makes the essay genuinely revealing. Balwit is not saying tech leaders literally think they are God. She is saying that by trying to build something of immense, even quasi-omnipotent power, they have wandered into the terrain religion has always occupied: creation, moral destiny, judgment, immortality, catastrophe, salvation. In her telling, many of these builders are not triumphant so much as spiritually unsettled. They are materialists working at the edge of questions materialism cannot comfortably answer. They are constructing systems of immense consequence while quietly sensing that the moral and spiritual frame surrounding them is too thin for the task.

From a Catholic perspective, that is both sobering and clarifying. Sobering, because it suggests the AI revolution is being driven in part by people who feel the stakes deeply yet do not possess a settled account of the good, of human nature, of humility, or of moral authority. Clarifying, because it helps explain why the Church’s voice may matter more than many Catholics still realize. Balwit’s essay is, in its own way, an admission that technical intelligence is not enough. The people closest to the machinery increasingly sense that they need wisdom, and perhaps even grace.

Her use of Chesterton is both fascinating and revealing. She sees around her a world of brilliant people who can reason powerfully about intelligence, prediction, and the future, but who remain haunted by questions reason alone does not settle. That is a profoundly Catholic insight. Man does not live by analysis alone. He needs wonder, meaning, worship, humility, and a moral horizon outside himself. Balwit recognizes that Silicon Valley has generated a kind of ersatz faith—one full of urgency and belief, but starved of transcendence.

That, finally, is why the essay is so significant for the broader AI debate. It shows that the question is no longer only what these systems can do. The deeper question is who the builders are becoming while they build them, and whether their moral and spiritual formation is adequate to the powers they are unleashing.

Balwit does not resolve that tension. But she names it with unusual honesty. And in doing so, she inadvertently strengthens the Catholic case: a civilization cannot safely develop quasi-civilizational technologies inside a spiritual vacuum and expect technique to make up the difference.

Tyler Cowen

Tyler Cowen’s tongue-in-cheek list of “seven ways” to avoid losing your job to AI is funny because it is true. His advice points in a clear direction: do work that is messy, embodied, relational, experimental, data-rich, and not easily reduced to a repeatable sequence at a screen. He says to look for “messy jobs,” to be wary of fully remote work, to become proficient with AI tools, to move toward fields like biomedicine, to run experiments, to gather data, and to seek more hands-on work.

In short: become more human where the machine is weakest, and more useful where the machine is strongest.

That is a sharper insight than the humor might suggest. Cowen’s list assumes that the safest future will belong less to the worker who simply sits at a terminal repeating standardized cognitive tasks, and more to the person who can move through the world, deal with unpredictability, test ideas in reality, gather fresh information, and interact with other human beings in concrete settings. That fits with his broader view that AI is more likely to create transitional disruption than permanent mass unemployment.

From a Catholic perspective, the striking thing is how much this sounds like a return to reality. If Cowen is even partly right, then the AI age may reward not disembodied cleverness alone, but presence, judgment, flexibility, and contact with the stubborn facts of the created world.

The joke hides a serious anthropological point: the more machines master abstract routine, the more valuable become those forms of work rooted in the body, in relationships, and in the unpredictability of real life.

There is something almost providentially ironic in that. A civilization that spent decades treating screen-based, standardized, white-collar work as the summit of prestige may now discover that many of the safer and more meaningful roles are those involving hands, presence, care, experimentation, and human adaptability.

Catholics should recognize the pattern. The dignity of work was never confined to the keyboard. The human person is not just a processor of information. He is an embodied being made for prudence, relationship, and action in the world.

So Cowen’s article works as comedy, but it also works as a warning. If you want to be less replaceable, do not become more machine-like. Become less so. Learn the tools, yes. But also move toward the kinds of labor where reality still pushes back, where other people still matter, and where intelligence must be joined to presence, character, and judgment. That is not just a strategy for surviving AI. It may also be a clue to recovering a saner vision of work.

Fr. Brendan McGuire

Fr. Brendan McGuire is, as I posted some time ago, a very important character in incipient but fruitful conversation between the Catholic church and those on the front line of artificial intelligence.

One of the most important observations he made in a conversation with Vatican News, is that the Church–tech conversation over AI is no longer a curiosity. It is becoming a real, structured, and increasingly mature dialogue. That matters because for too long the relationship between Silicon Valley and the Church could be dismissed as occasional symbolism: a conference here, a photo-op there, a few executives seeking moral cover. McGuire’s account suggests something deeper is now taking shape: a sustained conversation, nearly a decade in the making, rooted in listening, friendship, and a genuine search for wisdom.

That is what makes his testimony so significant. Fr. McGuire is not speaking as an outsider trying to comment on the tech world from a safe distance. He is a former engineer with a master’s degree in computer science and cybersecurity, now a priest in Silicon Valley, who has remained in close contact with leaders in the industry for years.

He describes how tech figures began coming to him not with triumphalism, but with anxiety—some worried by what they saw emerging, others asking what could be done. Out of that came listening sessions, deeper collaboration with Bishop Paul Tighe and the Dicastery for Culture and Education, the creation of the Institute for Technology, Ethics and Culture at Santa Clara University, and an expanding effort to accompany those shaping the future.

From a Catholic perspective, that is exactly the right development. The Church should neither flatter the industry nor retreat from it. She should do what Fr. McGuire describes: stay close enough to understand the real problems, morally serious enough to name them, and humble enough to become a trusted partner rather than a scolding spectator.

His language of a shared “search for wisdom” is especially important, because it points to what Silicon Valley often lacks most. The industry has no shortage of intelligence, capital, speed, or ambition. What it lacks is wisdom about the human person, moral limits, and the difference between what can be built and what should be built.

That is also why Pope Leo XIV’s Magnifica humanitas matters in this story. McGuire sees the encyclical not as the start of the conversation, but as a new intensification of a long journey. In other words, the Church is no longer merely reacting to AI from the outside. She is trying to speak into its development while there is still time, so that her contribution may be, as the article puts it, “more incisive and immediate.” That may prove decisive. A Catholic contribution to AI will matter only if it is not merely beautiful in principle, but present at the points where actual systems, companies, and norms are being formed.

According to an anonymous analysis of Pope Leo’s epochal encyclical Magnifica Humanitas, the document has used Artificial Intelligence “significantly” in the writing process, but concludes that the final product is completely human and totally consistent with the type of use of AI the Vatican recommends.

“The analysis identifies specific passages, structural patterns, and stylistic markers characteristic of large language model (LLM) output, while also noting evidence of substantive human theological direction and editing.”

As evidence of AI authorship, the report point out:

• Five consecutive sentences using identical “To speak of X means/calls for/obliges/requires Y” template

• Systematic coverage of all five Social Doctrine principles in order (common good, universal destination, subsidiarity, solidarity, justice)

• Each principle receives exactly one sentence—mechanically balanced

• No variation in syntax or rhetorical rhythm across the five iterations

• Pattern suggests prompting: “Apply each Social Doctrine principle to AI governance”

• Exhaustive enumeration: experiences, body, joy, pain, relationships, love, work, friendship, responsibility, moral conscience, good, evil, meaning, consequences—14+ distinct categories

• Systematic negation structure (”do not... do not... do not...”) repeated throughout

• Reads like a comprehensive response to prompt: “What capabilities does AI lack compared to humans?”

• The comprehensiveness itself is suspicious—human authors typically emphasize 3-4 key distinctions

“Magnifica Humanitas exhibits clear signatures of AI-assisted drafting in specific passages, particularly those featuring systematic enumeration, parallel structures, and comprehensive categorical coverage. However, the document’s theological architecture, bold institutional decisions, and obscure citations indicate substantial human direction and control.”

“The irony of an encyclical warning about AI being partially AI-drafted is notable but not necessarily hypocritical. The document itself calls for AI to be used as a tool under human direction, with humans maintaining responsibility for consequential decisions. If the hypothesized workflow is accurate, the Vatican has practiced what the encyclical preaches: using AI instrumentally while retaining human theological judgment and institutional accountability.”

Amarda Shehu, Vice President and Chief AI Officer at George Mason University, has published an essay that is one of the more penetrating things I have read on AI and the university because she refuses the usual evasions.

Shehu’s core claim is that AI is not simply adding pressure to higher education. It is exposing a dishonesty the university has lived with for decades: the university has spoken the language of formation, citizenship, and the life of the mind to donors and leaders, while selling students and families a credential tied to economic security. AI now threatens the credentialed performance that justified that bargain, especially at the entry level, and so the institution can no longer hide behind its old ambiguity.

That is exactly why her essay matters. She argues that large language models are beginning to absorb the kinds of digitizable tasks universities have long prepared students to perform: drafting, summarizing, coding, legal analysis, and other forms of credentialed entry-level knowledge work.

At the same time, she notes that tuition remains high, the labor market value of the degree is under pressure, public funding is constrained, and universities are responding with a vocabulary of “adaptation” that is far too small for the scale of the problem. Her point is that the institution is not facing a mere tool upgrade. It is facing a crisis of self-understanding.

From a Catholic perspective, that diagnosis is powerful because the Church was not involved in the birth of the university in order to create a credentialing machine. The university emerged from a civilizational conviction that truth is worth seeking, that knowledge belongs within a moral and metaphysical order, and that education is a work of formation, not just of certification.

Shehu’s argument, stripped to its essence, is that AI is forcing universities to answer whether they really still believe that. If the institution exists mainly to sell access to tasks now being automated, it is in serious trouble. If it exists to form judgment, relational depth, intellectual humility, endurance under uncertainty, and the capacity for slow understanding, then its mission remains not only defensible but indispensable. But, is it?

What makes her analysis especially strong is that she does not romanticize the old “holistic education” language. She sees that much of it has become aesthetic cover for a system that quietly monetized labor-market signaling.

She insists that the real unit of value is not the discipline label on the diploma but the formation a student actually undergoes. That is a brutal but necessary point. Some programs in “safe” fields may form almost nothing that resists automation, while some programs in supposedly doomed disciplines may form precisely the habits of mind and habits of relation that make a human being viable across a lifetime of technological disruption.

This is where a Catholic review should press even harder. Shehu is right that the university must redesign around formation rather than cling to the hedge of prestige plus credential. But Catholics should say plainly that formation is not just the cultivation of adaptable skills. It is the shaping of the person toward truth, virtue, prudence, solidarity, and a rightly ordered freedom. If universities respond to AI merely by shifting from “job training” to “resilience training,” they will still have missed the deepest point. The crisis is not only economic. It is anthropological.

So I would put it this way: Shehu has identified the challenge with unusual clarity. AI is not simply the latest disruption to higher education. It is the most dramatic challenge because it reveals that universities can no longer pretend to be two things at once. They cannot indefinitely sell a credential while speaking as if they were offering formation.

The Catholic answer is not to abandon the university, but to recover it: as a community of truth, apprenticeship, judgment, and human maturation, where the person is formed for more than the labor market precisely because the labor market is no longer stable enough to justify making it the soul of education.

Her final warning is the right one. If this redesign does not happen, the slow surrender will be paid for above all by students who can least afford it. That should hit us hard, because once again the burden of institutional dishonesty would fall on those with the least margin for error. In that sense, the AI challenge to the university is not merely technological. It is a justice question.

One of the more surprising early findings about AI in the workplace is that its use is not evenly distributed at all. According to a Financial Times poll of 4,000 workers in the United States and the United Kingdom, summarized by Tyler Cowen, AI adoption is heavily skewed toward the best-paid workers: more than 60 percent of top earners report using AI daily, compared with just 16 percent of lower earners. The same reporting also found that the heaviest users are not the youngest workers, but people in their 30s, and that men are using AI at higher rates than women.

That may be only a temporary snapshot. It is entirely possible that AI, because it is so widely accessible, will eventually spread much more evenly across the workforce. But if this pattern persists, it should worry us. A technology often described as democratizing could, in practice, end up reinforcing an older divide: those who know how to use the new tools, and those who do not; those whose jobs are amplified by AI, and those whose jobs are weakened by it; those who gain leverage, and those who quietly lose ground.

From a Catholic -and actually merely human- perspective, that would be a serious problem. A healthy society cannot be indifferent to the emergence of a new class divide built not simply on wealth, but on technological fluency. If AI becomes a force that disproportionately empowers the already educated, already skilled, and already well compensated, then the promise of accessibility will mask a deeper injustice.

The danger is not only that some workers will earn more than others. That has always been true. The danger is that AI may help create a new caste of technological “haves” and “have-nots,” even while presenting itself as universally available.

That is why Catholics should pay attention to this kind of data early. The question is not merely whether AI is efficient. It is whether its benefits are being distributed in a way consistent with human dignity, solidarity, and the common good. If the new tools are mostly strengthening the strong, then society will need more than optimism. It will need formation, access, training, and deliberate efforts to prevent a new inequality from hardening into place. Otherwise, AI may not abolish opportunity, but it may quietly ration it.

And that would be a very modern injustice: a revolution advertised as open to all, but most useful to those already ahead.

Reuters has been one of the news organizations at the forefront of analyzing the challenges AI presents to journalists, and this Reuters Institute piece by Marina Adami is another strong example. It shows that many in the industry are not simply panicking or retreating. They are trying to respond intelligently: building coalitions, developing shared technical standards, pursuing collective licensing, and even experimenting with AI products based on licensed content rather than theft.

These are serious and meritorious efforts because they aim at synergy rather than mere confrontation. They recognize that journalism cannot just scream at the storm. It has to find ways to survive inside the new environment.

The article lays out the problem clearly. Publishers fear not only that AI companies trained on their work without permission, but also that chatbots and AI-generated summaries are breaking the old internet bargain: platforms used to index content and send traffic back; now they increasingly consume the content and keep the user for themselves.

Reuters Institute research cited in the piece says only about one third of users consistently click through to source links on AI-generated search answers, while 28% rarely or never do. Similarweb data cited there also indicates that after Google launched AI Overviews, no-click searches increased and organic traffic to news publishers declined.

So the response described in the article is understandable. Groups such as SPUR are trying to build common standards and fairer licensing frameworks. Really Simple Licensing is trying to create machine-readable permissions and a collective rights system. Danish publishers, through DPCMO, have gone even further by negotiating together, partnering with technical tools to manage AI access, and suing when negotiation fails.

Others, like ProRata, are attempting a more collaborative model in which AI answers are built only from licensed content and revenue is shared through attribution. These are honest attempts to find viable forms of coexistence.

But my concern remains the same, and it is the word I keep returning to: convenience. The history of technology is ruthless on this point. When a tool offers users a dramatically easier, faster, and more frictionless experience, moral appeals and institutional defenses are rarely enough to stop it.

The Reuters piece itself all but admits this when several of its sources note that AI answers may simply be a better user experience than the old list of links. That is exactly why the threat is so serious. If generative AI keeps improving, getting faster, more accurate, more conversational, and more context-aware, it will continue moving dangerously closer to replacing not only parts of journalistic distribution, but in some cases the journalist’s visible public function.

From a Catholic perspective, that means the industry’s collaborative efforts deserve respect, but not illusion. They may win time, money, standards, and better terms. They may even help preserve some forms of journalism.

But they do not abolish the deeper civilizational pressure created by convenience. If the machine can answer quickly, smoothly, and cheaply, millions of people will choose it unless there remains something unmistakably human, trustworthy, and morally serious that they still desire from a journalist.

The challenge, then, is not only legal or commercial. It is anthropological. Journalism will survive only if it can still offer something that convenience alone cannot replace: witness, judgment, accountability, courage, and a human voice answerable to reality rather than merely trained on it.

Good luck out there colleagues!

In the latest post on his Substack newsletter, Richard Hanania, President of the Center for the Study of Partisanship and Ideology, stages a funny vanity test and half-turns it into comedy: can AI already write like me?

But what makes it unsettling is that the answer is not an obvious no. In his experiment, readers often struggled to distinguish his writing from AI-generated imitations, and in one case Claude came surprisingly close to fooling them. His conclusion is not that human writers disappear tomorrow, but that the old confidence that style, voice, and recognizable intelligence are uniquely ours is already being shaken.

That is where the post becomes more than amusing. Hanania argues that AI is getting good enough that many people can effectively outsource writing, that detection may become an arms race, and that in the future writing ability may no longer function as a reliable signal that someone is worth listening to.

He even suggests that credentials and preexisting reputation may matter more once original authorship can no longer be taken for granted. That is not just a clever internet problem. It is a cultural problem. It means one of the classic ways human beings have shown seriousness of mind—by wrestling thought into words—may lose part of its public meaning.

From a Catholic perspective, the funny part and the scary part are bound together. It is funny because AI exposes how much of what we call “voice” can be mimicked, stylized, and statistically reproduced. But it is scary because writing was never only about output.

Writing is a discipline of thought, an act of self-ordering, a way the person comes to know what he actually believes. If AI turns writing into something one merely commissions from a machine, then even excellent prose may conceal a thinner soul behind it. The danger is not only fake authorship. It is the weakening of the inner labor by which human beings seek truth.

And yet the Christian response should not be panic. Hanania himself notes something important: even in a world saturated with AI writing, readers will still want human beings. He thinks writing remains, at least in part, a parasocial relationship, and that as society grows wealthier, people will continue to prefer contact with minds like their own.

That instinct is more profound than he perhaps realizes. We are not only consumers of information. We are persons made for communion. We do not merely want correct sentences. We want to encounter someone through them.

So the Catholic takeaway is not that AI will eliminate writers, nor that it leaves writing untouched. It is that it may force us to recover what writing is actually for. If writing is merely the efficient packaging of ideas, then machines will do more and more of it. But if writing is also a human act of judgment, witness, struggle, and self-gift, then the machine can imitate much without replacing the thing itself.

That is why Hanania’s post is both hilarious and sobering. It shows that AI can mimic more of us than we expected. It also reminds us, perhaps against his own intention, that the truly human part of writing was never just style. It was the person behind it.

There is something melancholic in the thought that AI may be impoverishing language online. A recent study suggests that as AI-generated writing spreads across the internet, language itself may be becoming flatter, more repetitive, more sanitized, less alive. The complaint rings true. Anyone who spends enough time reading online can already sense the drift: smoother prose, more polished sentences, fewer surprises, less soul.

As a heavy user of AI for my own writing -especially since English is not my native language- I am guilty as charged.

And yet there is no point lying to ourselves. Cultural losses of this kind are usually unstoppable once a technology reaches a dramatic level of convenience.

The Church has seen this before. For centuries, the monastic scriptorium stood as a place of patience, discipline, beauty, and reverence. Monks copied texts by hand with care that was almost liturgical. Something precious was undeniably lost when the printing press made that world obsolete. A civilization of slow, crafted transmission gave way to speed, scale, and mass reproduction. But the press won because convenience won. And it did not win for trivial reasons. Books became cheaper, literacy spread, and access to knowledge broke out of older confines. A genuine loss was accompanied by a genuine gain.

I suspect AI and writing may prove to be something similar.

Yes, AI may degrade language in some ways. It may encourage shortcuts. It may reward formula over style, speed over struggle, fluency over depth. It may contribute to a culture in which fewer people learn how to write with force, precision, or beauty because the machine can do a passable version for them. That is not a small loss. Language matters because thought matters, and a culture that loses its feel for words often loses its feel for reality.

But IMHO, there is another side to this story, and it should not be dismissed. AI is also democratizing expression. Many people have something real to say but lack the technical writing skill to say it well. They have ideas, memories, convictions, insights, griefs, intuitions, and arguments trapped behind clumsy prose. AI can sometimes help bridge that gap. It can help a person express an idea that otherwise would have remained buried, not because the idea was empty, but because the craft was missing.

That is where a Catholic perspective should be both sober and hopeful. We should be sober enough to admit that convenience nearly always alters culture, and usually not without cost. The old scriptoria do not come back. The disciplines that technologies displace are rarely recovered in full. But we should also be hopeful enough to see that not every loss comes without compensation. A tool that weakens some forms of craftsmanship may also widen participation in the world of expression. It may make it easier for more human beings to speak.

So the question is not whether AI will preserve the old literary order intact. It will not. The question is whether we will still care enough about truth, beauty, and human depth to resist letting convenience become our only standard. If we do, then AI may serve as a tool without becoming a tyrant. It may help more people find words, even if it also tempts us to use them cheaply.

That, perhaps, is the strange consolation. The monastic scriptorium died, but the written word did not. Something like that may happen again. Language online may grow thinner in some places. But the ability to express an idea may also be placed in the hands of many who never had it before. And that, too, is no small thing.

My favorite economist, Tyler Cowen, makes a refreshingly sober point in his blogpost: AI is unlikely to produce mass unemployment, but that does not mean the transition will be painless. His argument is that the real trouble may come less from some final collapse of work than from temporary but serious frictions in the labor market. New jobs may arise in sectors such as energy and health care, but those are often heavily regulated and slow to scale.

Employers may also struggle to identify who is actually good at working with AI, which means hiring and matching could become clumsier before they become more efficient. And governments, trying to respond with fiscal policy or public investment, may be especially bad at sorting out who can thrive in the new environment.

That is an important correction to both utopianism and panic. Cowen is saying, in effect, that the labor market may suffer not because human work disappears, but because institutions are slow, confused, and badly adapted during the handoff.

That sounds right. It also means that workers may experience real instability even if the broader economy remains dynamic. There may be no “job apocalypse,” but there may still be layoffs, bad matches, stalled careers, and painful delays before new opportunities become visible and reachable.

From a Catholic perspective, that distinction matters. We should not exaggerate the case and declare that AI will simply abolish human labor. But neither should we hide behind reassuring aggregate forecasts while ordinary people absorb the shock.

Transitional disruption is still disruption. A worker who loses a job, cannot find the right fit, or watches institutions fumble their adaptation does not suffer in theory. He suffers in reality. That is why the moral burden here falls not only on markets, but on governments, employers, educators, and civil society.

So Cowen’s realism fits well with a Catholic view, provided one adds one more point: even if AI does not produce mass unemployment, it may still test our solidarity. The challenge is not only to trust that the economy will eventually re-sort itself. It is to make sure that the people trapped in the transition are not treated as expendable. The question is not just whether jobs will return in the long run. It is whether we will behave justly in the meantime.

AI may hit academic life, especially at the college level, far harder than many university leaders are willing to admit. In a conversation with Jay Caspian Kang, a staff writer at The New Yorker, Hollis Robbins, an American academic and essayist, argues that students will not only use AI to cheat, or that professors will need better policies. It is that much of higher education has already made itself vulnerable by becoming standardized, bureaucratic, and interchangeable.

Once courses become fungible, learning outcomes generic, and professors treated as replaceable delivery systems, AI does not appear as an alien threat. It appears as the logical next substitute.

That is where the OnlyFans line bites. Robbins is describing a future in which the credential may shift away from the institution and toward the individual professor: not the university as a coherent intellectual home, but the professor as a branded source of expertise, mentorship, and niche authority.

In her account, some of the future of higher education could look like “OnlyProfessors,” where students pay not for a broad institutional formation but for access to particular minds teaching at the edges of knowledge. She argues that institutions which survive AI will likely be leaner, stranger, and more differentiated, while much of today’s generic higher education may simply lose its reason to exist.

From a Catholic perspective, that diagnosis is both illuminating and incomplete. The Church was essential to the birth of the university precisely because the university was never meant to be a content vending machine. It was a community ordered to truth, a place where disciplines cohered, where masters formed disciples, and where knowledge was pursued within a larger vision of reality.

In that sense, Robbins may be right that the future could move away from mass standardization and back toward something closer to apprenticeship, mentorship, and intellectual discipleship. That part is not decadent. In some ways, it is pre-modern. It echoes an older educational logic in which formation came not mainly from interchangeable modules, but from sustained contact with living teachers.

But there is also a real danger. Discipleship can be noble; platformized intellectual celebrity can be corrosive. A university reduced to atomized expert-subscriptions would be a form of fragmentation, driven by market logic and personal branding.

Catholics should not mourn the collapse of empty bureaucratic sameness. But neither should it accept a future in which education becomes merely bespoke consumption, detached from a stable community of inquiry, moral formation, and shared pursuit of wisdom.

So perhaps the true alternative is not “OnlyFans or the university.” It is whether higher education will continue as hollow standardization until AI exposes its emptiness, or whether it will recover a deeper model that looks, in some respects, more like discipleship: teachers who really know something, students formed in living contact with them, and institutions confident enough to be distinctive instead of interchangeable.

As Kang implies, if AI forces that reckoning, it may not destroy the university so much as reveal how far many universities had already drifted from their original purpose.

A reasonable optimism about AI is not only compatible with a Catholic outlook. It should be. Ezra Klein’s essay on the New York Times is valuable precisely because it resists the fashionable fatalism that treats mass unemployment as inevitable and treats every new AI advance as proof that human work is about to collapse. He argues, plausibly, that history gives us reason to think technological revolutions often reconfigure labor more than they annihilate it, and that as automation expands, specifically human goods—attention, care, presence, trust, taste, judgment, relationship—may become more valuable, not less.

That basic instinct fits well with a Catholic realism. Christianity has never taught that human beings are doomed to be crushed by every new tool they create. It teaches, rather, that creation is intelligible, that human ingenuity can bear good fruit, and that labor, properly ordered, can be ennobling. So there is nothing un-Catholic about hoping that AI might free people from drudgery, expand prosperity, and even push society to rediscover forms of work and service that are more deeply human.

But Catholic optimism should probably be a little sterner than Klein’s. It must keep a closer eye on sin, power, and disorder. Wealth does not distribute itself justly. Technology does not automatically deepen solidarity. And the very relational capacities Klein thinks may become more valuable are, in fact, already being eroded by loneliness, isolation, and the technological habits of modern life. In that sense, his hope is reasonable, but it depends on preconditions that our society is actively weakening.

That is where Catholic optimism becomes both more cautious and more serious. We can believe AI will not necessarily bring a job apocalypse, while still insisting that it may wound workers, hollow out communities, and worsen inequality if left to market logic alone. We can believe that more “human” forms of work may flourish, while also recognizing that a culture formed by screens, fragmentation, and self-absorption may not be ready to supply the kind of mature human presence those jobs require.

So the Catholic position is neither panic nor naïveté. It is a realistic optimism: hopeful that AI can be integrated into a more human economy, but clear-eyed that this will not happen on its own. It will require moral formation, stronger families, better institutions, and a genuine defense of the dignity of work. In other words,

Klein is probably right to reject the apocalypse. But Catholics should add that the better future he gestures toward will only arrive if human beings still have the virtue to build it.

In a quite enlightening post, Derek Thompson’s basic point is that the deepest disagreement in the AI debate may be simpler than it looks: is AI a “normal” technology, or is it something historically unprecedented?

Drawing on Arvind Narayanan and Sayash Kapoor, he lays out the “normal technology” case: like electricity, the car, or the internet, AI may bring bubbles, monopolies, labor disruption, safety chaos, and regulatory catch-up without bringing civilizational collapse. Thompson notes that, so far, some evidence fits that view: several years after ChatGPT’s debut, growth and employment data still do not show the total social rupture many predicted, and even occupations once thought highly vulnerable have not vanished.

But Thompson does not stop there. He frames the real split as a clash between those who think AI will remain within the historical pattern of past technologies and those who believe it may cross into something radically different, especially if systems begin recursively improving themselves.

That is why the argument over whether AI is “normal” is not semantic at all. It shapes how people think about labor, regulation, national security, and even whether existing political institutions are adequate to govern what is coming.

From a Catholic perspective, that is exactly the right question to ask, though not in quite the same terms. The point is not merely whether AI is normal in the history of technology. The point is whether our response to it remains normal in the history of humanity.

Catholics should resist both cheap panic and cheap reassurance. If AI is “normal,” then it still requires moral discipline, political prudence, and a serious defense of workers, families, and the common good. And if it is not normal—if it is truly a threshold technology—then all the more reason to reject the lazy assumption that market incentives and technical elites will sort it out on their own.

So the value of Thompson’s post is that it clarifies the debate beneath the debate. The real fight is not just over capabilities, forecasts, or timelines. It is over analogy. Are we dealing with another electricity, or with something closer to a rupture in the human story? Catholics should be prepared for either possibility, but in both cases the task is the same: to insist that no technology, however “normal” or extraordinary, should be allowed to redefine the human person in its own image.

For me, the kind of news shared by Derek Thompson is not abstract. Pancreatic cancer is the disease to which I lost two beloved younger siblings. So when I read that AI may help detect this terrible cancer far earlier than the human eye can, I do not read it as one more tech headline. I read it as a reminder that artificial intelligence, for all its dangers, can also become an instrument of genuine mercy.

Derek Thompson’s piece points to something extraordinary: a Mayo Clinic team reported that an AI system called REDMOD identified subtle signs of future pancreatic cancer in old CT scans with significant accuracy, at a median lead time of 475 days before diagnosis. The article places that result alongside other recent advances, including promising drug results and personalized mRNA vaccine research, as part of a broader shift in the fight against one of the deadliest cancers.

From a Catholic perspective, this is the kind of development that should make us grateful, not fearful. If AI can help physicians see what even the best radiologists cannot yet see, and do so early enough to save lives, then it is serving something deeply human. It is serving the defense of life against suffering, delay, and death.

The moral test, as always, is whether technology remains ordered to the good of the person. Here, at least, one can glimpse what that good may look like: not a machine replacing love, but a tool helping love arrive in time.

That does not erase grief. It does not undo what families have already lost. But it does mean that in the middle of all the justified anxiety surrounding AI, there are also signs of hope serious enough to honor. And if one day this technology helps spare other brothers and sisters from the fate that took mine, that will not be a small victory. It will be a deeply human one.

Dean Ball’s “wave theory” matters because it is not really a theory about technology first. It is a theory about how human beings awaken to the significance of a technology.

In his essay, Ball argues that transformative ideas spread through society like waves: not everyone grasps their meaning at the same time, and when the wave finally reaches new groups, it often hits with much greater force. In the case of AI, the point is that recognition of its civilizational importance arrives unevenly, but once it does, politics, institutions, and public imagination can shift very quickly.

That is an important lens for understanding AI’s impact on humanity because it helps explain why public debate can look strangely delayed and then suddenly feverish. For a while, AI seems like a niche concern for engineers, founders, and a handful of policy people.

Then a capability jump, a public shock, or a visible crisis causes the wave to break much farther out, and suddenly business leaders, governments, and ordinary citizens realize they are not dealing with just another productivity tool. Ball’s claim is that awareness of advanced AI grows in amplitude as it diffuses, meaning later waves can be more socially and politically disruptive than earlier ones.

From a Catholic perspective, that insight is extremely useful. It means the moral question is not only what AI is, but when societies finally understand what it is doing.

Human beings and institutions are often late to recognize the full meaning of a revolution. By the time the wave reaches them, the technology may already be reshaping labor, security, culture, and even the way people imagine the human person. Ball uses the case of Anthropic’s unreleased model Mythos to argue that such moments can trigger a reset in policy and politics, precisely because more people are suddenly “hit by the wave” and forced to reckon with catastrophic risks and the likely role of the state.

That is why Ball’s framework matters for humanity as a whole. It suggests that AI’s impact will not be linear or smooth. It will come through shocks of recognition. One implication is that both denial and panic are temptations.

Ball explicitly argues that underrating frontier AI is a mistake, but so is overreaction, especially if fear hands excessive control to state actors without clear boundaries or public accountability. He calls instead for a structured, publicly legible, and clearly bounded role for government, especially around dangerous capabilities such as cyber vulnerability discovery.

For Catholics, the lesson is sharp. We should not wait for the wave to crash before speaking clearly about the human person, the common good, and the moral limits of technological power.

Although not explicitly Catholic, Michael Toscano’s basic argument could be quote well be adopted by Catholic concerned with where is AI going

His argument in an article in the New Atlantic that AI is not just another useful tool to be lightly supervised. It is part of a much larger technological project that tends to reorder family life, labor, culture, and even our understanding of the human person unless political authority pushes back. Drawing on Hannah Arendt’s critique of homo faber—man as fabricator and engineer—Toscano argues that modern society has grown accustomed to letting invention set the terms of life, and that AI now represents that tendency at full strength.

From there, his case for regulation is blunt: the state has both the right and the duty to say no, or at least not automatically yes, to technologies that threaten the common good. He points to earlier periods when political authorities restricted or redirected technologies to protect workers, families, or public order, and argues that modern governments have forgotten how to exercise that responsibility.

In his view, Big Tech will not restrain itself, because Silicon Valley is driven less by greed alone than by a deeper impulse to build whatever is “technically sweet” simply because it can be built.

The Catholic core of the argument is that technology must be subordinated to a human order, not the other way around. Toscano thinks strong regulation is necessary precisely because unchecked AI tends to treat human beings as raw material, replace human labor, and refashion relationships and institutions according to technological logic rather than moral ends.

Although I personally am less pessimistic, he makes a fair case, that is not merely prudential or economic. It is civilizational: if politics does not recover the courage to govern technology, then technology will go on governing us.

A recent Harvard-led study, published in Science and widely reported this week, found that OpenAI’s reasoning model outperformed physicians on several demanding diagnostic and management tasks.

In real emergency-department cases, the model achieved higher exact-or-near diagnostic accuracy than the doctors it was compared against, and the researchers argued that the results are strong enough to justify real clinical testing.

That is impressive, but does it mean medicine is about to become less human? Actually, it could be quite the opposite. The same reporting and expert commentary stress that these results came from retrospective testing, not from real-time bedside encounters, and that AI was not being evaluated on things such as reading distress, perceiving visual cues, building trust, or carrying the moral weight of accompanying a suffering person.

In other words, AI may prove better at certain forms of diagnostic reasoning without being better at the full act of healing.

From a Catholic perspective, that distinction matters enormously. If AI can shoulder more of the analytic and bureaucratic burden of medicine, then the vocation of the physician may actually become less robotic, not more.

Doctors could be freed, at least in principle, to spend less time functioning like overworked data processors and more time doing what no machine can truly do: listening, discerning, consoling, explaining, accompanying, and exercising prudent judgment in the presence of a vulnerable human being. That would not diminish medicine. It would restore it.

Of course, that better future is not automatic. Hospitals and health systems could just as easily use AI to chase throughput, cut staff, and push medicine even further into depersonalization. The technology itself does not decide which path we take. Human beings do.

So the real moral question is not whether AI can diagnose better in some settings. It is whether we will use that power to make medicine more humane or more mechanical. Catholics should be clear: if AI has a place in healthcare, it should be as a tool that strengthens the doctor-patient relationship, not one that quietly replaces it.

A recent Free Press piece by Ryan Hassan titled “The Filmmaker Using AI to Beat Hollywood“ makes plain what many still prefer not to admit: some of the most enthusiastic promoters of AI are not merely trying to improve creative work, but to blow up the old limits altogether.

In the article, filmmaker Charlie Curran celebrates AI precisely because it lets him produce fast, cheap, culturally reactive films at a speed Hollywood cannot match. He sees the collapse of the old system not as a tragedy, but as liberation: fewer costs, fewer gatekeepers, more output, more disruption.

That is exactly the spirit of technological maximalism now gaining ground: the conviction that if something can be accelerated, scaled, and democratized by AI, then it should be.

There is a real insight buried in that enthusiasm. AI may indeed lower barriers, widen access, and make room for creators who were once locked out by money, networks, and institutional control. But the moral problem goes deeper than the standard warning that we may be adopting a powerful technology too quickly.

The deeper problem is that speed itself is becoming a criterion of value, and efficiency is being mistaken for creativity. Once that happens, the question is no longer simply whether jobs will be lost, but whether human making itself will be redefined according to the logic of immediacy, scale, and endless production.

A Catholic response has to go beyond cautionary slogans. It must insist that creativity is not sacred merely because it is novel, nor human merely because a human prompted it. Art is bound up with discipline, patience, embodiment, apprenticeship, sacrifice, and a real encounter with reality. If AI becomes a tool that serves those things, it may have a legitimate place. But if it trains us to prize instant generation over craftsmanship, reaction over contemplation, and volume over depth, then it will not just disrupt creative industries. It will deform the human soul that creates.

So the real argument is not only that AI can move too fast. It is that a civilization can begin to love the wrong things in the name of innovation. And once that happens, warnings about “perils” are not enough. What is needed is a stronger account of what human creativity is for, and why no technological revolution, however dazzling, should be allowed to define that on its own.

A new National Bureau of Economic Research (NBER) working paper on generative AI and entrepreneurship points to a sobering but important reality: AI may not simply destroy jobs or create jobs in the abstract. It may do both at once, but not for the same people.

The authors find that startups with higher exposure to generative AI reduced employment quickly after ChatGPT, especially in junior and implementation roles, while productivity and financing outcomes improved.

At the same time, AI seems to be helping generate more startups, with more numerous but smaller investments and a rise in new firm formation.

That is a striking result. It suggests that AI is not only making companies leaner; it is also changing the kind of labor market those companies create. Entry-level workers appear especially vulnerable. The paper finds that employment reductions were concentrated among less senior employees, that about half of exiting workers faced roughly six-month delays in reemployment, and that many moved into lower-paying but less AI-exposed jobs.

But it is important to note that the employment generated by new firms largely offset overall losses in the aggregate, but with a shift toward managerial and senior roles.

From a Catholic perspective, this is exactly the kind of disruption that demands moral seriousness. It is good when technology raises productivity, lowers barriers to entrepreneurship, and helps create new firms. But it is not enough to celebrate “net neutrality” in total jobs if the actual burden falls hardest on younger workers trying to enter the economy.

A labor market that preserves opportunity only for the already skilled, already senior, or already established is not healthy simply because the aggregate numbers balance out. That is not how solidarity works.

So the warning here is plain. AI may expand dynamism at the top while narrowing entry at the bottom. If that is true, then the task is not to resist innovation as such, but to insist that innovation be accompanied by formation, retraining, and real pathways for younger workers whose first rung on the ladder is being kicked away.

Otherwise, AI will not abolish work. It will merely make access to dignified work more uneven. And Catholics should be among the first to say that this is not a side effect to be ignored, but a justice question to be confronted.

A recent essay by the Center of Digital Philosophy makes a serious philosophical claim: current AI may simulate many outward signs of mind, but that is not the same as instantiating consciousness; that is, actually bringing into being a real subject of experience -a unique human ability-.

The paper argues that computation, by itself, is an abstract description imposed on physical processes, not yet a sufficient account of lived awareness. On that view, AI can imitate the map without becoming the territory.

That is a useful corrective at a moment when some developers and commentators are tempted to speak as if generative AI were already approaching human consciousness.

Even Anthropic’s own research on “introspection” is much more limited than the hype suggests: it reports only some degree of introspective awareness in current Claude models, says that this capacity is highly unreliable and context-dependent, and explicitly states that the results do not tell us whether Claude or any other AI system is conscious, nor whether it introspects in the same way or to the same extent that humans do.

From a Catholic perspective, that matters. Human consciousness is not just clever output, internal monitoring, or persuasive linguistic performance. The human person is an embodied creature with intellect, will, moral agency, and a spiritual destiny that no artifact can simply replicate by becoming more statistically fluent. AI, as a human artifact, will never rival God’s creation in its full ontological depth.

That said, the future challenge may not be a machine that truly becomes humanly conscious, but one that operates so convincingly as if it were conscious that societies begin to treat it as such. That is where confusion, moral slippage, and anthropological disorder can begin. The paper’s distinction between simulation and instantiation helps clarify exactly why that line matters.

So Catholics should resist two errors at once: naïvely declaring today’s generative AI conscious, and complacently assuming the question is therefore irrelevant.

It is relevant precisely because systems that are not conscious may still be treated as if they were—socially, legally, emotionally, even liturgically or spiritually by the confused. The task ahead is to defend a robust account of the human person before simulation becomes culturally indistinguishable from presence.