By Solveig Strauss. Cover installation view of Three Approaches to Synthetic Entities 3 “Monologue” display at the Mori Art Museum in Tokyo, Japan 2025. Courtesy of Diemut Strebe.
Where is the art world headed as artificial intelligence reshapes the very notion of creativity? It’s a question artists and institutions are wrestling with, but few probe its fault lines as boldly as Diemut Strebe. The German-American artist, now in her second residency at MIT’s Center for Art, Science & Technology (CAST), operates in the charged space where art, science and philosophy collide. Collaborating with leading researchers such as Robert Langer and Daniela Rus, Strebe uses large language models and emerging forms of AI agency not merely as tools but as conceptual provocations — ways to test how far technology can push, unsettle or even reinvent human authorship. Her work suggests that the future of art won’t simply be about machines creating images, but about the evolving choreography between intuition, code and the unknown.
Can Machines Really Be Creative? Exploring AI as a Partner
Many of your works use AI not just as a tool but almost as a creative partner. From where you stand, is AI capable of genuine creativity — or are we still looking at human imagination refracted through an algorithmic mirror?
An AI machine, specifically LLMs, cannot make art. It extracts patterns and predicts the next token, but has no clue of art, thinking, or feeling. The human condition that informs art substantially plays no role in the matrices of an AI machine. Beethoven’s 2nd symphony is not a pattern of notes. It is a vision about music and the transcendence of human mental states condensed into an aesthetic experience. Nevertheless, AI can create creative results – if you train it correctly, it can come up with amazing ideas, but these are the result of a training process and a probability distribution, not the extraction of meaning. It doesn’t know it had a great idea, except for a pure reinforcement.
Teaching Machines to Understand Human Emotion and Aesthetics
You speak about “teaching” the machines you work with. What does it mean for a machine to learn things that are intrinsically human, such as aesthetics or emotion?
It means that I prompt the machine with a role it has to play, then review the outcomes and fix any loopholes. For my show in Tokyo, the machine had to act like a human — which it is not — so it had to lie constantly. And it lied convincingly, even better than humans do! Still, these matrices do not actually process emotions, neither the positive nor the negative ones.
When Authenticity Meets Algorithms: Is AI Art Less Real?
AI inherently strips us of creativity. How would you respond to critics who view this type of art as less “authentic” compared to handmade works?
Authenticity lies in the artistic concept. It can also be present in the execution, but it doesn’t have to be. The idea of “cutting off the artist’s hand” relates to a famous Duchampian quote on the subject. Jeff Koons never touched his famous sculptures, which were produced by a German company. Maurizio Cattelan never sculpted his kneeling Hitler.
The Future of Art and Humanity in an AI-Driven World
Finally, what does the future of mankind look like to you with AI’s growing presence — one of coexistence, or one where it surpasses us?
There’s a high chance it will be very challenging, due to humans being replaced by machines. Another major concern is that we may never fully crack the black box of AI and could end up ruled by systems we cannot even understand. Once a machine can comprehend and calculate how parts relate to a whole, it could optimize for goals that are not aligned with human preferences. With the exception of the mother-child relationship, the weaker force rarely controls the stronger one. In addition, people often gravitate toward the lowest-energy states to achieve their goals. Only a few humans will persist — because they do what they do regardless of circumstances: they create.
