The question about AI and art is not whether the output is good. It is what happens to a person who no longer struggles to produce it

The argument about artificial intelligence and art has settled into two entrenched positions, and neither is quite right.

One holds that AI-generated work is derivative by nature, assembled from what already exists, and therefore not art at all. This is emotionally satisfying and increasingly difficult to defend on the evidence. Some generated work is genuinely striking, and audiences who do not know its origin frequently cannot tell.

The other holds that this is simply the latest tool in a long line, no different from the camera, the synthesiser, or the word processor, and that objections are the familiar complaint of those displaced by change. This has the advantage of historical humility and the disadvantage of dissolving a real question by analogy.

The more useful question sets output aside entirely. Ask instead what making does to the person who makes.

The Formation Hidden in the Struggle

Anyone who has learned an art form knows that the difficulty is not incidental to it.

A person learning to draw spends months producing work they find embarrassing. In that time something is happening that has little to do with the drawings: they are learning to see. They notice how light falls across a face, how a shoulder actually joins a neck, how much of what they assumed they saw was a symbol they carried rather than a thing in front of them. The bad drawings are the mechanism by which the seeing is acquired.

A writer discovers what they think by trying to say it and finding the sentence will not hold. The struggle is not an obstacle to expressing a pre-existing thought. It is how the thought becomes definite. Anyone who writes seriously knows the experience of arriving somewhere they had not intended to go because the paragraph refused to work.

A musician practising a difficult passage is not only training their hands. They are learning that something impossible in January is possible in March, which is a lesson about persistence that transfers to everything else they will attempt.

None of this appears in the finished work. It appears in the person.

This is why the comparison to previous tools is not quite apt. The camera did not remove the photographer’s need to see; it changed what seeing required. The synthesiser did not remove the musician’s need to hear. Generative systems are different in kind because they can complete the task without the maker passing through the process. The output arrives. The formation does not.

The Theological Stake

The Christian tradition holds that the human person is made in the image of a God who creates, and that human making participates, distantly and truly, in something divine. This is not a claim that art is a religious activity. It is a claim about what the capacity to make reveals about what a person is.

Human creativity has three features worth noticing, and they are precisely the features generative systems do not share.

It is embodied. A painter’s hand has limits, and the limits shape the work. A singer’s voice is a particular voice with a particular history. What is made carries the mark of the body that made it, and the mark is not an imperfection to be engineered away.

It is costly. Making takes time from a finite life. The cost is why a made thing is a gift when it is given, and why the work of a person’s hands has meant something in every culture that has existed.

It is answerable. A person who makes something stands behind it. They can be asked what they meant, can be praised, can be blamed, can be changed by how the work is received. A generated image has no one behind it in this sense. There is a prompt, and a system, and no one who intended precisely this.

A civilisation that produces enormous quantities of images and text without anyone having undergone the formation of making them has not simply become more efficient at culture. It has broken the connection between what is made and the persons who make.

What Was Taken

There is also a straightforward matter of justice that gets lost in the more philosophical arguments.

These systems learned from the work of living artists, writers, photographers and musicians. That work was used, almost universally without consent, without compensation, and often without acknowledgement. Models can now produce work in the manner of specific living artists, some of whom find themselves competing against a facility built from their own life’s output.

Catholic social teaching has been unambiguous for over a century that a worker is entitled to the fruits of their labour and that value created by a person’s work cannot simply be appropriated by whoever has the means to extract it. Rerum Novarum said it about industrial labour. The principle does not change because the extraction is computational.

Whatever one concludes about the aesthetics, the training question is a labour question, and it has been answered so far by the party with the most resources rather than by anyone weighing the claim.

Where This Is Genuinely Good

An honest account has to include the other side, because it exists.

A person with a severe motor disability can now produce visual work they could not otherwise make. Someone who cannot afford an illustrator can give their story a cover. A person with vivid inner images and no training can externalise something and see it. Composers can hear an arrangement before commissioning musicians. Small parishes and community organisations can produce materials that previously required a budget they never had.

There is real democratisation here, and dismissing it is easier for people who already had access to the means of making.

The distinction that matters is between using these tools as part of a practice and using them instead of one. An artist who generates variations to think through a composition is still doing the work of judgement, which is much of what art is. A person who generates an image and posts it has made a choice, but has not undergone anything.

Both are legitimate uses. They are not the same activity, and confusing them is where the argument usually goes wrong.

What Is Worth Protecting

Not the art market, and not the professional status of artists, though both matter to the people whose livelihoods they are.

What is worth protecting is the human experience of making: the months of bad drawings, the sentence that will not come, the passage practised until it yields. These are formative, and they are formative for everyone, not only for those who become professionals. A child who learns an instrument acquires something that persists whether or not they ever perform.

The risk is not that AI produces bad art. It is that a generation grows up with no reason to endure the difficulty, and therefore never acquires what the difficulty produces. Not the skill, which is replaceable, but the patience, the attention, the tolerance for one’s own inadequacy, and the discovery that persistence changes what is possible.

Those capacities are not about art. They are about being a person capable of anything hard.

The Practical Response

For families and schools, the response is not prohibition, which will not work and would not be right. It is ensuring that young people have experiences of making things slowly with their own hands and minds, and that these are valued rather than treated as inefficiency.

For parishes and communities, it means continuing to commission human work, sing with human voices, and remember that liturgy has always been made by particular people in particular places.

For anyone who makes anything, it means understanding that the practice is not only a means of production. It is a means of becoming, and that part is not automatable, because it was never about the output.

The finished work was always the by-product. The person was the point.