Generative Artificial Intelligence (AI) became widely known with the arrival of ChatGPT on the consumer market in November 2022. Numerous tools capable of producing text, images, music or even videos have emerged and transformed uses in both personal and professional spheres. But behind the media frenzy, how well do we really understand how it works? Are proprietary services the only option?

Many training courses have emerged to introduce professionals to these technologies, often from a generalist perspective and focused on office or text-based applications.

The objective of this course is to develop a detailed understanding of how generative AI technologies work, to broaden perspectives, and to reclaim creative processes. To achieve this, it offers an alternative to dominant proprietary services by exploring open-source tools — the only approach that enables real control over the tool, transparency of methods, and freedom of use.

This course will focus on ComfyUI, a modular (node-based) graphical interface based on Stable Diffusion, which allows the creation, visualisation and execution of complex creative protocols (workflows) for images, text, sound and video while maintaining precise control over each step of the creative process.

Audience

This practical course, interspersed with theoretical insights, is aimed at individuals involved in content production (text, image, video, audio) who wish to gain a critical and in-depth understanding of how generative AI works and free themselves from the limitations of proprietary tools, which can often be restrictive in terms of aesthetics and customisation.

Some familiarity with proprietary generative AI tools (Midjourney, Krea, ChatGPT) is helpful but not required.

Prerequisites

• Basic IT skills: familiarity with computers, the internet, and digital tools.
• A PC with an NVIDIA graphics card OR a Mac M1/M2/M3/M4, and at least 50 GB of available storage.

Desktop computers will be provided for those interested who do not have a laptop meeting the technical requirements.

Learning outcomes

• Gain an understanding of how artificial intelligence works by exploring open-source generative tools through hands-on practice.
• Discover and experiment with open-source alternatives such as Stable Diffusion, and identify the philosophical, technical, and economic differences between these open tools and mainstream proprietary solutions (Midjourney, OpenAI, Adobe).
• Develop the ability to use ComfyUI and design your own workflow.
• Engage in collective reflection on the ethical issues, technical limitations, and responsibilities associated with the use of generative AI.

Course format

Practical workshop.

Dates and hours

Dates: the dates for the next session will be announced at a later date.
Hours: 09:00–12:30 and 13:30–17:00

Detailed programme

Taught by