A quotation from Mario Klingemann as an epigraph:
“For me, AI is just one tool in a long history of tools that was inevitably going to be used for artistic purposes. I would say that I use AI as a tool, and that the works I create with it are mine, rather than the result of a collaboration — just as I would not call a hammer or a piano a ‘co-author’.”
These days, you often hear people say: “Neural networks will replace us all — artists, designers and people in other professions too.” It can be frightening, especially for someone who has spent a lifetime drawing and putting their heart and soul into their work. Now something has entered the picture that can “paint a picture” in a few seconds.
But is it really so frightening? I am not a technology enthusiast, nor do I believe in progress at any cost. I love and value art and craft. I visit exhibitions with great interest, have already collected works by contemporary artists and am always looking out for more. I also experiment with neural networks because I am curious. That is where it all began.
This article is simply a reflection.
We have been here before — more than once
Looking back, we can see that art has always changed alongside its tools. People first drew with charcoal on cave walls. Then came paints, though not the kind we know today: they were complex mixtures made from natural pigments. They had to be mixed by hand, were difficult to store and were used in workshops.
When paint in tubes was invented, artists could take it with them for the first time — and went outdoors. Impressionism followed. One tool, and a whole new aesthetic.
The same is true of brushes. One kind of brush gives you one technique. Another produces a different stroke, a different kind of painting. Yet a new brush has never made the previous one cease to exist. They continued to coexist.
The arrival of the camera was hardly uneventful either. I am sure nineteenth-century artists worried: a portrait could now be made in a second, rather than through hours of sitting. Until then, painting had striven to be as precise and realistic as possible. Now there was a click — and it was done.
But what happened? Painting did not disappear. On the contrary, it began searching for something new. Impressionism, Cubism, Expressionism and Surrealism emerged — forms in which accuracy mattered less than conveying a feeling, an emotion, the artist’s way of seeing.
The camera became a distinct tool rather than an enemy. Analogue film gave rise to an entire culture. Then digital cameras arrived, and the arguments began again: “That isn’t real.” Once again, nothing disappeared. There was simply more.
We now have cameras in our phones, but that has not made everyone a photographic artist. Paints sitting on a shelf do not make someone an artist either. A tool is only a tool. What matters is the idea, and how to give it artistic meaning and expression.
Neural networks are more than “one button”
AI is not a magic button labelled “make it beautiful”. Creating work that resonates, that fits a style, an idea and an intention, takes effort. You need to know a great deal, try many things, understand how different neural networks work, and bring together styles, materials and meanings. Sometimes it takes months to arrive at a single visual technique.
I think many people imagine they can write, “Make me a painting that will sell for millions at auction,” and a masterpiece will be ready. But it does not work that way. It is like buying an expensive brush and expecting it to do all the painting itself.
Take Mario Klingemann, for example. He is one of the best-known artists working with AI. He began experimenting back in the 2010s, wrote his own algorithms and was there at the beginnings of AI art. His first work sold at auction appeared in 2019. Almost a decade of work.
Artists who already work with AI — and successfully
There are artists today, recognised within the professional art community, who openly say, “I use AI in my practice.” Their works are shown in museums, galleries and biennials.
Mario Klingemann (Germany) works in the field of AI art. His works have been exhibited at the Centre Pompidou and MoMA, and sold at Sotheby’s, where an artwork fetched £40,000.
Refik Anadol (USA/Turkey) creates large-scale installations. In 2023, his project “Unsupervised” was at MoMA.
Sougwen Chung (Canada) draws with an AI-powered robot. She explicitly presents her projects as a dialogue between the artist and AI. Her work has been exhibited at the New Museum and HOFA Gallery.
Anna Ridler (UK) works with neural networks using datasets she assembles by hand. Her works have been shown at the V&A and MoMA.
Egor Kraft (Russia) makes sculptures using AI. His work has been exhibited at ZKM, Anna Nova Gallery and MMOMA.
They are fellow explorers, rather than competitors to other artists. Their brush happens to be a neural network.
There is one thought I particularly want to convey: neural networks are simply a new tool. Neither the first nor the last. You can experiment with AI as you do with photography, light or collage. In every case, however, it is the person who decides what they want to say. It is the person who brings the idea. AI does not create meaning. It can only help express it.
Artists have existed, exist now and will continue to exist. After the invention of the camera, after digital cameras, and now that neural networks have arrived. Some will continue to paint with a brush. Others will try something different. All of this will coexist.
Neural networks are not a threat. They expand the palette, like a new brush, a new paint or a new filter. If you are curious, try them. It is in your hands.
