2026 Online Workshop Series
Ethical Creative AI Tools
Drawing on two decades of research and creation at the Metacreation Lab for Creative AI, this workshop series presents ethical tools for computer-assisted creative work. The series covers music composition (both audio and MIDI) and visual generation (Autolume and Model Bending in ComfyUI), both for offline asset creation, installations and live performance.
For each of the tools presented, we will detail their motivations, demo the tool, and give examples of real-world use. These workshops highlight the very existence of tools that are not falling into the pitfalls of corporate “big data” AI, and introduce small data, model crafting and model personalization as central to the ethical use of AI tools in artistic practices.
These workshops aim to offer a critical, yet balanced, introduction to a creative toolset demonstrating how AI can become a controllable, adaptable, and artistically valuable creative tool.
Day 1
Music with MIDI-GPT
August 19, 2026, 10 am-12 pm PDT
Discover MIDI-GPT, a generative AI system developed by the Metacreation Lab for computer-assisted music composition. In this hands-on workshop, participants will explore how to generate, continue, and rework multi-track MIDI compositions while controlling instruments, musical style, note density, polyphony, and duration. No advanced AI knowledge is required. Just come with an interest in experimenting with new approaches to music creation.
Day 2
Autolume
August 20, 2026, 10 am-1 pm PDT
Autolume is a no-coding, user-friendly, visual synthesizer that leverages the artistic potential of GANs (Generative Adversarial Networks) without requiring technical skills. It streamlines the GAN workflow, including dataset preparation, model training, and real-time generation, making it accessible for non-technical users. Our workshop welcomes researchers and artists across all visual disciplines and technical backgrounds. Participants will be guided through a comprehensive system demo, learn how to leverage Autolume for training models with their unique datasets, and harness Autolume’s interactive capabilities, including audio-reactivity.
Day 3
Day 4
Fragmenta
August 21, 2026, 10 am-1 pm PDT
Fragmenta is an open-source app that brings the complete generative AI text-to-audio pipeline to musicians: intuitive dataset creation, LoRA training, generation, audio editing, and live performance. Developed by Metacreation Lab’s visiting researcher Misagh Azimi, and built on Stable Audio 3, Fragmenta is designed for all musicians, and especially for experimental music and sonic arts practitioners, giving them the ability to shape the latent space with their own audio and musical data, no coding required. It reflects the small-data, model-bending, and artist-first approaches to AI.
Model Bending with ComfyUI
August 22, 2026, 10 am-12 pm PDT
Model Bending is a creative, experimental technique inspired by digital circuit bending and glitch art, applied directly to generative AI systems. Instead of relying solely on text prompts, it intercepts a model's internal activations and layer parameters during generation to explore visual variations, create abstract results, or reveal how concepts are represented under the hood. By discovering and tweaking control points hidden in the model, you can unlock glitch aesthetics and unique visual styles that standard prompting simply cannot produce.