SLEEK: Your work has evolved from figurative painting through a creative crisis to collaboration with AI and science. How did that come about?
Roman Lipski: It developed naturally. It was about finding a personal expression for what was happening in my mind and to communicate that. Without Berlin, I would never have become an artist. I was 19, in a new city in the early 90s, and decided to become a painter without having ever painted before. A lot of it was improvised and self-taught.
I painted for over 20 years, then I realised I was repeating myself. I had ran out of stories. I felt like a robot. The collectors were satisfied, but for me it was a creative death because I was no longer experimenting.
This desire for change led to a crisis around 2014. I wanted to remain a painter, reinvent myself, and abandon figuration. In 2015, I met data scientist Florian Dohmann. We developed the first AI models for painting without replacing the painter. This resulted in the AI Muse.
When I changed my style, I lost all my collectors and was left broke, but I was happy because I had achieved my goal. Collaborating with scientists led to quantum art. Today, I work with a team and new patrons in my studio and exhibition space in Kreuzberg. I am free to experiment. My encounter with AI and science has expanded my horizons, but it has not replaced me. Through crisis, technology, and collaboration, I have returned to painting: freer, more open, and more successful than before.
S: You use artificial intelligence as a tool in your artistic practice. How does this dialogue with the machine shape your work process in the studio?
RL: Every time I use technology, it’s for support. I use it to help me find new solutions or to inspire me. A dialogue develops between me as a human being and the machine. Artificial intelligence and quantum mechanics are tools in my studio. The AI Muse is like a personal AI agent and was trained with my material. The algorithms analysed my painting—colours, shapes, brush structures—and reassembled them. This resulted in variations of my visual language and the effect was like a revelation to me.
I recognised myself in them; suddenly, I saw solutions. The AI didn’t invent anything new, but rather recombined my material. I saw my motifs in an abstract, pointillist, surreal way—possibilities that were already there. These digital images were not finished works of art, but a catalogue of possibilities.
The important thing is that as soon as I’ve learned something, I switch the AI off again and work analogue. Machines can speed up processes, but art is not created algorithmically. Artistic decisions arise in the act of doing, in the act of painting.
S: AI also led you to quantum art—how is that reflected in your artworks?
RL: When it comes to the results of AI, they were consistent with my work because they were based on my material. I deliberately kept the digital images small on the screen. I wasn’t tempted to print them large and show them as my own works. That helped me to use them only as inspiration. The artwork only comes into being when I translate these impulses into physical painting. AI also led me to quantum art and brought me into contact with quantum physicists. The idea of combining my AI Muse with concepts from quantum physics immediately fascinated me. I am less interested in the technology than in the aesthetic and conceptual expansion of pictorial spaces.
Today, this is reflected in my paintings not as an illustration of science, but as an attitude: open spaces, vibrant color structures, and nonlinear compositions. The digital results are not the artwork, but the starting point. The artwork remains material, color, and gesture.
All Photography by BASTIAN THIERY