The Art of Non-Prompting Workshop
Most AI art advice is about control — better prompts, tighter steering, more predictable output. This workshop flips that. We'll explore "non-prompting": treating AI as an unpredictable collaborator rather than a tool you direct, using minimal prompting and maximum freedom for the machine, with your role shifting from operator to curator. You'll see how this approach plays out in real work, then try it yourself — first with image and video, then with sound driving image.
who it's for:
Filmmakers, visual artists, and AI-curious creatives who want to move past "prompt engineering" — anyone feeling boxed in by trying to control AI output, or curious what happens when you let go of the wheel. Also useful for festival programmers and educators thinking about how AI collaboration actually works in practice, not just in theory.
what you'll take away:
A working framework for "non-prompting" — minimal input, maximum machine freedom, curatorial control
Hands-on practice with non-prompting across image and video
A direct look at how sound shapes perception of image (and vice versa)
Your own short video piece, built from a shared audio prompt
A take-home guide: how to make (AI) art without fear
speaker info:
ArielKo is an Israel-based visual artist and designer working across AI art, branding, illustration, and large-scale site installations. His AI art practice grew out of "non-prompting" — deliberately minimal prompts (often just two or three words) used as a reaction against the trend of increasingly elaborate, control-heavy prompting, letting the AI's own unpredictability lead the work. That approach runs through recent projects like The Human Race, Field Notes, and collaborations tied to Peter Gabriel's music, alongside a long-running design career spanning branding and campaign work for Teva, Google, and Intel, and permanent public installations including Tel Aviv's light rail stations and several memorial and park sites in Israel. He teaches this same non-prompting approach hands-on, guiding participants through building original AI-driven work in real time rather than lecturing about technique.