Can AI truly be creative?

Somehow, I was disappointed. But why? The quality of the illustration hadn’t changed. Nevertheless, the image suddenly felt trivial. This moment between admiration and disillusionment sums up where we currently stand with AI. Namely, at a crossroads. And that reveals something fundamental about people and their relationship to creativity, art, and what we can expect from AI.rnrnAs a branding and design studio, we navigate this tension every day. Every day we see new AI-generated designs, logos, or campaigns. Agencies boast about AI integration. Freelancers fear for their livelihoods. And in the middle of it all is a pretty big question: What is creativity these days, and can programmed software actually be creative?

If we want to understand whether generative AI can be creative, we must first ask ourselves what creativity actually is. And it is more complicated than you might think. The scientific consensus on the definition of creativity is as follows: »Creativity is the ability to use one’s imagination to develop novel and valuable ideas or works.«

In our opinion, creativity requires two things above all else: originality and usefulness. Something new that also works. According to this definition, generative AI would indeed be creative. It can spit out original images that are certainly useful. Case closed? Not quite.

Creativity researchers and experts argue that this is not enough and falls short. It only describes the result, but not the process of creation. It’s like judging a writer solely on the basis of the finished novel, without asking whether it was written by the author themselves or copied from someone else. True creativity requires two additional dimensions: intentionality and authenticity.

Intentionality – the conscious decision

A creator makes conscious decisions. Selecting something, discarding it, experimenting, starting over. Always with intention.

Take Wolfgang Amadeus Mozart, for example. He deliberately composed his »Musikalischer Spaß« to be bad, as a satire on the amateurish composers of his time. Another example from the world of design is David Carson, with his chaotic hierarchies, destroyed grids, and illegible text. When he designed layouts in the 1990s that broke every rule, it was not incompetence, but a conscious rebellion against the Swiss school. Both examples show that the quality of the work also lies in the process and the way it was created. A generative AI may know the rules, but it cannot decide for itself to break them, and it cannot weigh up for itself when chaos is enriching and when it is just chaos.

Authenticity – the self in the work

Authenticity means that the work is irreplaceable. It can only be created by this person, at this time, with these experiences. This is because a work usually expresses something unique about the person who created it.

This includes perspective, experiences, or the self. Why do we recognize a Picasso at first glance? Not necessarily because he perfected a particular new painting technique. Rather, it is his visual identity. His obsessions, his breaks, his worldview, and his entire life flowed into every brushstroke. A good example of this is one of his most famous paintings: »Guernica«. It was created in 1937 in response to the destruction of the Spanish city of the same name by an air raid.

Generative AI has no experiences of its own, has not experienced highs and lows, and has no self to express. It merely combines what others felt and created. The result may be impressive, but it is far from authentic. But even if we understand these theoretical dimensions, why does it still feel so wrong when we are deceived?

Imagine you go to a concert. A pianist in a smart tailcoat sits at the Steinway grand piano. His hands dance impressively fast across the keys, the music is breathtakingly beautiful. After the encore, you find out that the piano is just a facade and the music came from a loudspeaker. The artist was only pretending to play. How does that make you feel? Somehow like you’ve been cheated, right?rnrnThis is where a fascinating psychological effect comes into play: we humans evaluate performances differently when we know how much effort went into them.

A schoolchild’s pencil drawing of a crooked heart moves us more than a perfectly printed greeting card design because we can sense the effort and intention behind it. A personalized handwritten letter is special because someone took the time to think carefully about its content and perhaps even started over several times due to spelling mistakes. A painted portrait is special because it requires a great deal of skill and time. Today, supposedly creative individuals enter a prompt, press »Enter«, sit back, and proudly present the result. They pretend to have created something that was actually produced by an algorithmic process, composed of millions of other people’s works that were fed into a database without permission.

For generative AI, these things are no problem. A pencil drawing by a schoolchild? A personal text? A portrait in the style of an oil painting? The task doesn’t matter. Everything is implemented without any apparent effort, and the result doesn’t really impress us because we know that it took no effort at all.

Let’s take a quick look at our industry. When desktop publishing emerged in the 1980s, established designers said, »This is the end of the design profession!«. Anyone could now create a layout using PageMaker. But if you wanted to do really good work back then, you still had to understand everything about typography, type area, and hierarchies.

The software was just a tool; in-depth knowledge of design remained essential. Then Photoshop came along, and with it, another outcry. But to work really well with Photoshop and master it, you need a lot of know-how and practice. Then came the internet and, once again, numerous changes. Suddenly, anyone could build a website. But there is a world of difference between a template website and a well-designed digital brand experience. Every disruptive innovation in human history was initially met with great skepticism. History has taught us one thing: new technologies change jobs. However, industries as a whole remain intact. AI experts also expect AI to improve human capabilities, but not replace them entirely. Nevertheless, we are now facing a paradigm shift because we must recognize that generative AI is something radically new.

The use of generative AI is comparable to a repetitive briefing. You tell the software what it should do, preferably in as much detail as possible, and it delivers a result. A little more blue here, a different word there, or a completely new structure for a text. Overall, it’s an iterative process until you get a result you’re happy with.

Is that reprehensible in itself? Not really. But you should be honest and transparent with yourself and, above all, with your customers about who did what work here. In this case, generative AI is the actual creator. This leads to an exciting paradigm shift. You go from being the client and creator back to being the contractor.

Perhaps the problem also lies in the fact that we do not yet have a suitable definition for the results of AI. At the moment, we are trying to describe something radically new with old words. The outputs of generative AI could perhaps be summarized under the term »pseudo-creativity«. They are simply a particularly good imitation.

But why don’t we generally refer to the results of generative AI as »artificial creativity«? Just as we use the term »artificial intelligence« (it is not intelligent in the human sense, but it simulates intelligence), we should refer to »artificial creativity« (it is not creative in the human sense, but it simulates creativity).

Artificial creativity can be original and effective. It can be surprising and useful. But it lacks essential dimensions: intentionality, authenticity, intrinsic motivation, independent problem-finding, and aesthetic responsibility. In short: everything that characterizes human creativity.

So, can AI be creative? As is so often the case, our answer is: it depends. In this case, it depends on how you define creativity. If creativity simply means “creating something new that works,” then yes, AI can be creative. But if creativity means consciously designing something, expressing it authentically, and taking responsibility for aesthetic decisions, then no. AI will never be creative in this sense.

But the more interesting question now is: Does generative AI threaten human creativity? Yes, artificial intelligence is already better than us humans in many disciplines, and it will surpass us in even more areas in the future. But that doesn’t mean human creativity will disappear. Quite the contrary. A mistake many people are making right now is to believe that the aesthetics and tonality of generative AI cannot be recognized.

And the more generic AI mediocrity floods the world, the stronger the desire for unique works that embody human know-how and sensitivity will become. As long as this is the case, human creativity will never disappear. It will only become rarer and therefore even more valuable. And we are convinced that human creativity will always reign supreme.