Artistic Practices - Artificial intelligence

HUM-322 / 2 crédits

Enseignant(s): Jankauskas Vytautas, Udvari Tibor

Langue: Anglais

Remark: Une seule inscription à un cours SHS+MGT autorisée. En cas d'inscriptions multiples elles seront toutes supprimées sans notification

Summary

Artificial Intelligence (AI) is changing how we imagine stories, experiences, and interactions. Rather than treating AI as a threat to existing creative practices, we will explore how different AI tools can open new ways of thinking, making, and prototyping.

Content

Working in groups, you will use ready-made art-and-design-oriented AI models to turn ideas into frugal interactive prototypes. You will learn to iterate and refine these prototypes through design ethnography and user testing, to then build a final project.

The aim of the course is to teach basic AI techniques and iterative methods media artists and interaction designers use to prototype and effectively communicate new ideas.

Course material is centered around demos and hands-on exercises with easy-to-use AI models, covering the following principal themes:

Models trained on custom small datasets (Teachable Machine)
Computer vision and body, facial, gesture tracking (ml5.js, MediaPipe)
Language models (LLMs) built into the web browser (e.g. Gemini in Chrome)

Sessions include art and design examples that use AI in interesting ways, while allowing space for more critical reflection on AI's impact on creative practices and methods.

Throughout the course, groups build 2-3 screen- or object-based prototypes, be it a music instrument, a mini game or toy, a provocative interactive experience, etc. The strongest prototype grows into a more refined final project.

During the first session, a theme is given by the tutors, to guide the prototypes. Prototypes are judged by the quality of created experience (what they make someone feel, notice or question). Feedback sessions with the tutors discuss strengths and feasibility of ideas. When the final project starts taking shape, groups conduct user tests (observation, short interviews) and iterate designs from the findings.

Since students work with technologies that do not yet come with fixed rules of use, nor with familiar visual languages, the goal is to build confidence in prototyping by learning to decide what a prototype should (not) do, and by exposing it to users early-on.

Keywords

Artificial intelligence, media art, interaction design, practical course, prototyping, user testing, design ethnography.

Learning Prerequisites

Important concepts to start the course

No preliminary skills required. The course is intended for students new to media art and design, coding, and artificial intelligence.

Learning Outcomes

Produce small interactive prototypes using art-and-design-oriented AI models
Implement a small creative model based on a custom dataset
Conduct a user test (observation, short interviews) and revise the prototype from their findings
Create a coherent, visually sound presentation

Transversal skills

Demonstrate a capacity for creativity.
Take feedback (critique) and respond in an appropriate manner.
Evaluate one's own performance in the team, receive and respond appropriately to feedback.
Negotiate effectively within the group.

Teaching methods

A hands-on creative studio: each experiment starts from a short brief and demonstration, with supervised in-class time for experimenting and building projects in groups. Work-in-progress is regularly discussed with tutors and other groups. AI assistants may be used to help with coding tasks.

Expected student activities

Students work in groups of three and are expected to actively participate in the class: building prototypes and exploring AI tools, giving and receiving critique, as well as keeping a simple logbook of their process (a template will be provided). Groups are formed in the first weeks and stay together for the semester. Some work to polish prototypes may be required outside class hours.

Deliverables (per group):

3 interactive prototypes (see timetable below)
User testing documentation (week 13; see timetable)
1 advanced prototype (chosen from among the 3 prototypes; week 14; see timetable)
Process logbook (week 14; see timetable)

Timetable:

Week 1 (Feb 23) â€“ Introduction - Round of presentations
Week 2 (Mar 2) â€“ AI tools overview; vibe coding with AI

Week 3 (Mar 9) â€“ Tutorial and brief #1: Custom small datasets (Teachable Machines)
Week 4 (Mar 16) â€“ Practical studio and critique
Week 5 (Mar 23) â€“ Practical studio and critique

Week 6 (Apr 6) â€“ Tutorial and brief #2: Computer vision and body tracking (ml5.js)
Week 7 (Apr 13) â€“ Practical studio and critique
Week 8 (Apr 20) â€“ Practical studio and critique

Week 9 (Apr 27) â€“ Tutorial and brief #3: LLM in the web browser
Week 10 (May 4) â€“ Practical studio and critique
Week 11 (May 11) â€“ Practical studio and critique

Week 12 (May 18) â€“ Final project; design research methods
Week 13 (May 25) â€“ User testing and production
Week 14 (Jun 2) â€“ Final presentation

Assessment methods

A strong project is judged by its ability to communicate on its own. Someone trying the prototype for the first time should be able to understand it and use it without elaborate additional explanations.

20%: proactive participation (assessed individually)
40%: prototypes (made in groups; assessed per group)
30%: final project (developed, user-tested, presented to the class; assessed per group)
10%: documentation (assessed per group)

Code quality is not part of the grade.

Resources

Bibliography

Golan Levin & Tega Brain, Code as Creative Medium (MIT Press, 2021).
Jennifer Walshe, 13 Ways of Looking at AI, Art & Music (Unsound, 2025)
Lauren Lee McCarthy, Casey Reas & Ben Fry, Getting Started with p5.js (Maker Media, 2015)
Kate Crawford, Atlas of AI (Yale University Press, 2021)
Going further (critical AI): Joy Buolamwini / Coded Bias; Hannah Davis, "A Dataset is a Worldview"; Ruha Benjamin, Race After Technology
Spike ePaper â€“ Issue 77: Field Guide to AI (2024)

Projects:

https://www.whatbeatsrock.com/

https://driesdepoorter.be/happysadchat/

https://www.andreasrefsgaard.dk/projects/face2wikipedia/

https://github.com/trishume/LastSecondSlides

Online project collections:

https://mlart.co

https://experiments.withgoogle.com/collection/ai

https://www.creativeapplications.net/

Tutor websites:

Vytas Jankauskas : https://vjnks.com/

Tibor Udvari: https://tiborudvari.com/