At the Ecole Nationale Supérieure d'Architecture et de Paysage de Bordeaux (ENSAP Bordeaux), a group of pioneering undergrads are discovering how AI tools can transform architectural creativity. Led by Professors Fabienne Darricau, Jean-Marc Emy, and Vincent Witomski, the course, called "Artificial Intelligence: When Representation Becomes Project," turns generative AI into a hands-on design ally.

Why AI in architecture?

Some people see AI as a limitless source of inspiration. Others worry it might replace genuine human ingenuity. This course navigates that tension by challenging students to embrace generative AI for exploration without losing their own creative spark. Here, AI is not the solution. It is more like a co-creator that nudges you toward unexpected ideas.

A peek into the course structure

The studio runs like an experimental workshop. Students explore both the theory and the practice of AI in design.

1. Kickoff and background. A quick look at how digital tools have shaped architecture so far, from CAD to BIM to parametric design, and an introduction to the buzz around AI.

2. Hands-on with Floyo and ChatGPT. Students dive into Floyo, which runs open source diffusion models, alongside ChatGPT. They try prompt engineering, iteration, and collaborative critique. The main goal is to see how AI-generated visuals can spark new architectural concepts, especially around green architecture.

3. Critical feedback. Every so often, everyone gathers for group critiques. How do these AI-driven visuals align with real sustainable design principles? How do they shape, or misshape, our ideas of green building?

4. Show and tell. Students share final image sets, references, and design proposals. The emphasis is on explaining why certain AI outputs make sense, and where human intuition still matters most.