Skip to main content

Concordia and SAT researchers create open-source, digital creation tools for artists

‘The SAT is not simply presenting AI as a closed black box. They are building ways for artists and creators to understand, test, and shape these systems directly’
April 17, 2026
|
An indoor music event with a DJ and a backdrop of blue light making geometric shapes
Jean-Michaël Celerier: “We are continuously improving the tools for embedded development for media arts.” | Photo courtesy of the Society for Arts and Technology

“Artists are really on the lookout for tools that grant them the power of AI models while keeping their data safe and local,” says Jean-Michaël Celerier, technical development director at Montreal’s Society for Arts and Technology (SAT) and an alumnus of the postdoctoral research program at Concordia’s Faculty of Fine Arts.

That shared concern is propelling a growing collaboration between Concordia and the SAT, connecting research, creation and tool development. For his part, Celerier is helping lead efforts to design tools that combine generative AI, computer vision and immersive technologies while addressing ethical and artistic questions around AI in art.

Diptych image of a young man with long, dark hair and a beard Jean-Michaël Celerier. | Photos courtesy of Jean-Michaël Celerier

Giving artists the wheel

SAT recently released a set of open-source, artist-centred tools designed to support the creative process while maintaining artists’ control over data, authorship and workflows. Rather than relying on opaque commercial platforms, this approach allows artists to host their work on their own servers.

“We are continuously improving the tools for embedded development for media arts, and trying to empower artists with ownership of tools that they could run on extremely small systems,” Celerier says.

At Concordia, his research focused on ossia, a free intermedia tool he explored in collaboration with the Critical Practices and Materiality Research Chair. The project is led by Alice Jarry, associate professor of design and computation arts.

“The goal was to integrate ossia with embedded devices and systems and work on low-power alternative to tools that students use,” Celerier says.

One outcome is Ballets Résiduels, a research-creation project and polysensory performance and installation he created with artist Brice Ammar-Khodja. Running on ossia in a Montreal park, it shows how powerful creative systems can operate on minimal hardware.

From a Raspberry Pi Zero 2 W — an ultra-small, single-board computer — the project maps itself in real time using GPS, processes sound, generates haptic feedback and diffuses smells, all while running for hours on a phone battery.

Portable, trainable systems

This emphasis on lightweight systems reflects a broader shift at the SAT. Celerier and the SAT research team is now focused almost exclusively on open-source, local-first AI.

“For the next two years, we will have research creation projects exploring how artists use open-source tools to explore new artistic ideas,” says Edu Meneses, research director at the SAT.

“Open-source tools let us build 100 per cent locally, giving us control over the data, including how AI generates it. That's a key point of discussion where we can effectively use AI tailored specifically to artistic practice,” he says.

Trainable AI allows artists to achieve outcomes in their own style, he adds, with benefits including portability, longevity and improved access for historically marginalized communities.

Support for art students

The collaboration is also expected to expand at Concordia, with plans to involve students more directly through the AI and Creativity Summer Intensive. The studio-based course will be led by Christian Beltrami, the university’s inaugural AI expert-in-residence.

Beltrami says the goal is to help students engage with AI as a flexible, expressive medium.

“We want students to explore AI as a multimodal creative system. That means looking at how body movement, gesture, voice, style, and even emotion can become inputs that drive outputs such as images, music, prompts, 3D worlds and narrative.”

He adds that the approach emphasizes experimentation and transparency.

“SAT is not simply presenting AI as a closed black box. They are building ways for artists to physically navigate the AI's latent space, moving away from generic, predictable outputs to uncover completely new, unexplored artistic vocabularies.”

That shift, he notes, also reframes creative practice.

“We are not just asking what AI can generate, but how the body, space and performance can act as a creative interface. Those possibilities are not theoretical: they can be explored through tools that connect gesture, pose, sound, image and space in real time.”

The AI and Creativity Summer Intensive runs from May 11 to June 1, 2026. Registration is now open. To learn more, or to apply, visit the course web page.


Discover the
Department of Design and Computation Arts at Concordia.

 



Back to top

© Concordia University