What We Owe Our Own Taste

We spend a great deal of time developing ourselves. We learn new skills, pursue education, exercise our bodies, develop professional expertise, and try to become better at the things that matter to us. Yet we rarely speak about developing our capacity to appreciate things. Taste is usually treated as something we simply have. I like this. You like that. There is not much more to say.

But perhaps there should be.

If some works of art are better than others, and if some experiences are richer or more rewarding than others, then it seems reasonable to ask whether we should cultivate our ability to recognize and appreciate them. Do we owe it to ourselves to develop better taste?

The question is especially interesting in an age when artificial intelligence can curate our aesthetic experiences for us. An AI system can recommend music, films, books, clothing, restaurants, artwork, and countless other things based on what we have previously enjoyed. This can expand our horizons. It can also create a strange possibility: we may become increasingly good at consuming things we like without becoming any better at judging them ourselves.

The difference matters.

Taste Has to Be Earned

Consider Friedrich Nietzsche’s changing relationship with Richard Wagner. Nietzsche initially regarded Wagner with extraordinary enthusiasm. He devoted years to Wagner’s music and thought, wrote extensively about him, and became one of his most important philosophical interpreters. Eventually, he turned sharply against him.

In one of his late notebooks, Nietzsche writes that he had changed his mind about Wagner, and that he won’t grant just anyone the right to make the same judgment. At first this can sound arrogant. What he means is that there is a process behind a conclusion like that. Nietzsche had gone through phases of enthusiasm and admiration before he arrived at his reversal. There are no predetermined stages, but it is not a casual opinion, and it cannot be reached without some process.

There is an important difference between saying, “I don’t like Wagner,” and arriving at that conclusion after years of listening to him, studying his music, comparing him with other composers, thinking about his aims, and understanding what initially attracted you to his work.

The first is a preference. The second is an achievement of judgment.

This does not mean that familiarity automatically makes a judgment correct. We can spend years liking something for bad reasons. Nor does expertise require that we eventually admire what experts admire. Nietzsche’s example is instructive precisely because his extensive engagement led him away from Wagner rather than toward him.

What matters is that the judgment was earned through experience. Developing taste requires more than accumulating preferences. It requires attention, comparison, memory, reflection, and a willingness to remain with things long enough to discover what they have to offer. Sometimes it requires learning to appreciate something we initially dislike. Sometimes it requires discovering that something we once admired no longer satisfies us.

Taste develops when our encounters change our capacity for future encounters.

Better Taste and Personal Taste

This raises a harder question. What would make one person’s taste better than another’s in the first place?

Philosopher Jerrold Levinson takes up this problem in his essay “Artistic Worth and Personal Taste.” Levinson begins with two claims that seem difficult to reconcile. First, there are better and worse works of art, and the same is true of aesthetic objects generally. Second, our personal tastes matter to us. We do not want to become machines that merely recognize artistic quality. We want our aesthetic lives to have some connection to who we are.

The tension is real. If there are genuinely better works of art, and if experiencing them can enrich our lives, then we have reason to improve ourselves as aesthetic appreciators. We might reasonably want to become more perceptive, more informed, more capable of recognizing artistic value.

But there is a problem with taking this too far.

Suppose I could transform myself into a perfect aesthetic judge, with excellent taste in every domain: the finest paintings, the greatest novels, the most accomplished music, no longer limited by my own undeveloped preferences. Would I have improved my aesthetic life?

Perhaps. But Levinson points out that something else might be lost. Our aesthetic preferences are part of our individuality and our personal history. If developing taste meant progressively eliminating everything distinctive about our responses until we converged with an ideal critic, we might become better judges while becoming less distinctive people. Levinson calls this the “paradox of aesthetic perfectionism”: we have reason to improve our aesthetic capacities, and also reason to preserve our individual aesthetic personalities.

Developing taste, then, cannot simply mean learning to like what someone, or something, tells us is best.

What AI Can Curate, It Cannot Cultivate

This is where our contemporary situation becomes especially interesting. AI is good at curation. Give it enough information about your preferences, and it can recommend things likely to appeal to you. It can identify patterns you might not recognize yourself. It can introduce you to an artist, composer, filmmaker, designer, or author you might otherwise never encounter. There is nothing inherently wrong with this. Good curation can be an important part of developing taste.

The problem arises when curation replaces cultivation.

Imagine I tell an AI that I enjoy a particular kind of music. It recommends twenty albums I am likely to enjoy. I listen to several and like most of them. I ask for more. The system learns from my responses and becomes increasingly accurate. I may now have an excellent personalized collection of music. But what have I learned? Perhaps very little.

We can receive increasingly precise answers to the question, “What will I probably enjoy?” We may not become any better at answering the harder question: “Why is this good?”

Those are different abilities.

AI can help us find things. It can even help us understand them. But if we always let it determine what deserves our attention, we risk losing some of the experiences through which taste itself develops. The danger is not that AI will recommend bad art. The subtler danger is that it will recommend good matches for our existing preferences, so good that we rarely have to develop preferences of our own.

The Value of Getting It Wrong

Taste requires failure.

We need to encounter things we dislike, misunderstand, initially dismiss, and later come to appreciate. We need to discover that our first judgment was incomplete, and encounter works that resist our expectations. This is part of what makes aesthetic development different from simple consumption.

Suppose you hear a piece of music and immediately decide it is not for you. You could move on to something more familiar, and an algorithm can make that transition almost frictionless. But perhaps the piece has qualities that only become apparent after repeated listening. Perhaps you need to learn its conventions, or hear what the composer is doing before you can appreciate why it works.

Sometimes the thing you initially dislike becomes something you love. Other times, you return to something you once admired and discover you no longer care for it. Both experiences are valuable. They teach you something about the object, but they also teach you something about yourself.

This is one reason the development of taste requires more than exposure. It requires engagement: enough freedom to wander, enough patience to remain with difficult things, and enough independence to disagree with the recommendations we receive. An algorithm can tell us what resembles what we already like. It cannot substitute for the experience of discovering that our previous preferences were too narrow.

Developing Taste Without Losing Ourselves

Levinson’s worry was that perfecting our taste might mean converging on an ideal critic and losing what is distinctive about us. None of this means we should try to eliminate personal taste. Quite the opposite.

The goal of aesthetic cultivation should not be to produce people who all like the same things. Nor should “good taste” simply mean familiarity with an established canon, or the ability to recognize whatever cultural institutions currently regard as important.

A cultivated taste can remain deeply personal. The difference is that a cultivated taste is responsive. It notices distinctions that an undeveloped taste misses. It can encounter unfamiliar things without immediately reducing them to familiar categories. It can recognize excellence even when that excellence does not correspond to one’s immediate preferences. And it can change.

This is why Nietzsche’s example is so useful. His rejection of Wagner was not the abandonment of taste for the sake of personal preference. It was itself the product of a long aesthetic history, a judgment made more discriminating because his experience had become more extensive and reflective.

Developing taste does not require us to become an ideal critic in Levinson’s sense. We do not need to trade our individuality for a universal aesthetic standard.

Perhaps what we owe ourselves is something more modest, and more important: the opportunity to become more perceptive participants in our own aesthetic lives.

What Do We Owe Ourselves?

So, do we owe it to ourselves to develop better taste?

I think the answer is yes, provided we understand “better” carefully. We do not owe ourselves a particular list of paintings, books, composers, designers, or films we are supposed to admire. We do not need to become cultural snobs, and we certainly do not need to surrender our preferences to an algorithm or an expert.

We owe ourselves the cultivation of the capacities that make genuine taste possible: learning to pay attention, exposing ourselves to things that challenge our existing preferences, becoming willing to change our minds, learning enough about a work to understand what it is trying to accomplish, comparing, remembering, reflecting, and sometimes simply spending more time with something before deciding what we think of it.

AI can help with some of this. Used well, it can introduce us to unfamiliar territory, provide information that makes deeper engagement possible, and become a genuine tool for cultivating taste. But it should remain a tool. The final judgment should still belong to us.

There is a difference between having a machine curate an aesthetic life for us and using a machine to help us develop the capacity to curate our own. The first gives us increasingly personalized experiences. The second gives us something more valuable: a growing ability to decide for ourselves what is worth experiencing.

Perhaps that is what it means to develop taste. Not learning to like the right things, but becoming capable of appreciating more of what is worth appreciating.

And perhaps that is something we really do owe ourselves.