Artists Navigate the AI Frontier (Keynotes)

Creative Showcase Session - June 3, 2025

Five Swiss artists at Panoramai's creative showcase revealed compelling tensions between artistic ambition and technological reality, demonstrating varied strategies for maintaining creative agency in an age of artificial intelligence.

From AI Enthusiasm to Critical Distance

Paul Petrino shared his evolution from AI evangelism to thoughtful skepticism through "3038.50.97," a science fiction book created with writer Anton de Macedo using Midjourney's third-generation model. The numerical title reflects their process: « 3038 is the number of images we generated just to create the first maquette, and 50.97 is the hours we spent on mid-journey. »

They deliberately stayed on Midjourney's version 3, finding its unpredictable artifacts more compelling than polished newer iterations. « The images on the right, the images are a little bit boring. And something on the left was really fascinating for us, » Petrino explained, comparing early AI outputs with current commercial-focused models.

Their workflow generated thousands of images organized through Discord channels corresponding to fictional biomes. Three subsequent exhibitions transformed these static images into immersive experiences, culminating in manually deconstructing 100 books and weaving them into a single new volume—deliberately inverting AI's data synthesis process.

« We realized that for us, it was crazy that we could be inspired by AI, but create something without AI, and that we were super happy, » Petrino reflected. His current disillusionment stems from AI's commercial optimization: « Right now the new models are made to sell stuff... for us that are trying to do something in fine arts, it's not interesting anymore. »

Robotics as Poetic Expression

Léa Peyrere, representing the Claire+Léa design duo, transforms industrial robotics into organic, breathing sculptures through their ANIMA series. Working with ETH Zurich's democratized Pathos platform, they create forms that respond to human presence rather than demonstrate technical prowess.

« The nice thing is, as I said, this technology was democratized enough to be used by someone not being able to program, so ourselves, » Peyrere noted. Their installations consistently surprise audiences who assume the floating, breathing forms are AI-generated imagery rather than physical robots.

Extending into educational robotics at EPFL, Peyrere collaborates with students to reimagine technical fields through aesthetic approaches. « What drives me now is this kind of educational aspects... trying to bridge the gap between people that feel distant with technology. »

Her latest project, ANIMA 5, uses machine learning to translate music into robotic movement, allowing an ABB arm to "dance" to compositions like Ravel's Bolero through collaboration with EPFL's REASSIST lab.

Computational Minimalism

Boris presented a contrasting philosophy of technological simplicity, deliberately countering AI's complexity through geometric animations and interactive installations. Following a personal health crisis, he developed counting-based visualizations that assign colors to sequential numbers—a meditative practice yielding unexpectedly rich results.

« I try to do the opposite of what artificial intelligence pushes us to do, that is, I try to simplify things rather than to complexify them, » Boris explained. His browser-based interactions explore emotional connections through minimal computational gestures, finding meaning through constraint rather than possibility.

Hybrid Photography

Mathieu Bernard-Reymond brought decades of photographic experience to AI experiments, having witnessed analog-to-digital transition around 2000. His Disco Diffusion work generates "missing images" from family albums—moments between documented memories.

« I realized their ability to work a bit like I had the impression that my memory was working, » Bernard-Reymond observed. His process combines textual prompts from Swiss author Charles Ferdinand Ramuz with rock texture photographs to structure AI-generated landscapes.

His ambitious portable AI camera project generates images locally without internet connectivity, bridging computational photography with traditional field work. Though he admits: « I didn't expect to be so confused... I realized that maybe I was a photographer because I don't like to talk a lot. »

Artificial Divination and Technological Magic

Raphael Lutz and Meret Vatslavik concluded with "Divination," an interactive installation that deliberately subverts rational AI expectations by positioning it as an oracle. This critical design project explores societal assumptions about artificial intelligence.

« What place do we give to Artificial Intelligence? And why do we give it a blind spot? We have a blind faith in the reasoning of Artificial Intelligence, » Vatslavik questioned. Their machine combines mystical aesthetics with large language models, asking users to interpret symbols before receiving personalized predictions.

Testing with 150 users across Swiss cantons revealed fascinating cultural variations. « La Chaux-de-Fonds wanted to play. They didn't want to understand, they wanted to play, » Vatslavik observed. The project's evolution toward more provocative responses deliberately tested AI interaction ethics, with one user becoming so disturbed they refused their printed prediction.

The Democratization Paradox

Throughout presentations, artists emphasized how complex tools—from robotic programming to machine learning—had become accessible to non-technical creators. Yet this accessibility came with trade-offs between capability and control.

Each presentation revealed different strategies for maintaining creative agency within increasingly automated systems. From Petrino's journey to post-AI creation, through Peyrere's poetic robotics and Boris's computational minimalism, to Bernard-Reymond's hybrid processes and Lutz-Vatslavik's critical mystification, these artists demonstrate that meaningful engagement with AI requires wrestling with its implications rather than accepting its outputs uncritically.

Their collective work suggests AI's most lasting creative impact may not be the content it generates, but how it reshapes fundamental questions about authorship, intentionality, and human-machine collaboration in artistic practice.