A friend shared an interview with Caterina Moruzzi, a philosopher exploring how people and AIs work together, and what that relationship does to us as humans. As I listened to it, I became increasingly aware that we need to define what “creativity” is more precisely in light of what AI can do for us. She is working with artists and musicians, so the line between the artistic and the creative is ambiguous in this interview. Let’s take a closer look at the difference.
I want to first establish that something artistic, beautiful, or tasteful is not necessarily creative; and something creative is not necessarily artistic. The key difference is that creativity is more objective than we often assume. Here are some examples.
Let’s say we need to build a bridge but the location is not conducive to building any solid structure. Someone could come up with a beautiful design for it, but it cannot be built at that location. Another person comes up with a solution that allows us to overcome the limitation of the location. The former, as beautiful as it may be, would not be considered “creative” because the result does not achieve our objective, and this objective is definable. Creativity, at least in the way we often use the term, is about making unexpected connections to solve a definable problem. In many cases, we can even measure it in terms of profit, efficiency, or productivity. Without this socially agreeable value, we are not likely to use the term “creative.”
On the other hand, let’s say you have an empty wall you want to cover with something beautiful. You commission an artist and she creates a painting that elevates the entire space, but a blunt friend walks in and says, “I think it’s ugly, and countless artists have made similar paintings. This is not creative or original.” Be that as it may, you personally still think it’s beautiful even if nobody else does. To you, it is still “artistic” because art is in the eye of the beholder.
Now, you might complain that my definitions are too narrow. Fair enough; I’m narrowing them on purpose because AI forces us to separate what is objectively solvable from what is subjectively meaningful. In everyday conversation, we mix these terms freely (creativity and artistry blur together), but that vagueness makes it impossible to see what AI is actually encroaching on. The two concepts lie on a spectrum of objective and subjective values, but for the sake of this argument, I need to pull them apart. I’m not suggesting we always use them this way; I’m doing it here so we can see what part of “creativity” AI can surpass and what part it can’t.
I would assume that AI will soon surpass us in “creativity” as defined above. AI or machine learning models are easy to train when the goal can be objectively defined. Recognizing letters and numbers or identifying flowers were among the first problems solved with AI because the correct answers were easy to obtain.
In other words, AI will be superior to us with any problems that have objectively correct answers; it is only a matter of time. So, creativity, after all, isn’t all that “human” in the sense that its superiority is independent of any of us. I would not mind outsourcing it to AI entirely. What is more human is the quality of being “artistic.”
I’m currently working on a project with an artist where our code generates artworks from randomized parameters and has him approve or disapprove them based on his personal taste, what he thinks is beautiful, regardless of what anyone else thinks. The machine learning model we built can now predict with roughly 80% accuracy whether he would approve another randomly generated work.
Now imagine ten years from now, when AI is exponentially smarter: Would this model be superior to or more “artistic” than the artist? I hope you can see that this is a nonsensical question. The very nature of art resists objective comparison. As soon as the model deviates from the artist’s taste, it simply becomes a different taste, not a superior one, and it becomes less useful to him because its predictive accuracy declines.
Just as Duchamp tried to demonstrate by putting a urinal in a gallery, art is about communicating what you think is beautiful; it is this assertion that is more meaningful than the work itself. AI can certainly help artists create artworks, but it can never be superior (or inferior, for that matter) to humans because there is no basis for comparison.
This also pertains to products we don’t normally consider artistic. Take a computer application. Now that AI can help any of us create our own apps, it will soon be pointless to buy apps designed for millions of others. Your needs and preferences are unique; why put up with sifting through features you don’t need? Why not have AI build exactly what you need, like ordering a custom-tailored suit? Although we don’t normally use the word “artistic” in this context, the subjective dimension of UI/UX design is structurally the same: it reflects personal preferences and tradeoffs that cannot be resolved objectively. That is the “artistic” aspect of UI/UX.
As you try to design one, you will realize that reality forces you to make many compromises, not because your budget is limited, but because reality has conflicting forces. Sometimes we want life to be predictable and routine so we can take care of things efficiently by following the same pattern over and over, but other times we crave freedom and flexibility. No matter how large your budget is, you cannot design an app that accommodates both simultaneously. You could design two separate interfaces, one rigid and one loose, but most of the time, we want something in between. If we designed a thousand different interfaces, managing them would create a problem of its own.
As we confront our own desires, tastes, and preferences, we quickly notice that they pull in opposing directions. The same structural tension appears in our social reality, which also makes contradictory demands on us. From a scientific or engineering mindset, such conflicts look like problems waiting for solutions. Artists, however, recognize that many of these tensions are fundamentally irresolvable: confidence vs insecurity, kindness vs resentment, solitude vs connection, freedom vs responsibility, honesty vs kindness, fear vs desire, love vs hate. Each pair contains truths that cannot be collapsed into a single answer, nor repressed without consequence. Art is the process of metabolizing these contradictions rather than resolving them. Each person must discover their own way of living with them; there is no universal formula.
In this sense, as AI advances and allows us to get exactly what we want, we will confront our own contradictions. Once we find and embrace them, we won’t care what other people say about them. The app you designed works for you and that is all that matters. Whether other people find it useful is secondary; maybe they do, maybe they don’t.
It is this artistic dimension, not the creative one, that will be more meaningful in our relationship to AI.
I will email you when I post a new article.
