Procedural Translation of A.I. Generated Motives into Architectural Design

Artificial Intelligence generated images have become widely available. Artists have been using A.I.s such as Stable Diffusion in creating creative interpretations based on textual prompts. There is an ongoing debate whether arts produced using A.I can even qualify as “art”. Artists are questioning the validity of the technology. But it is still a great tool, using machine-learning to aid the creative process. Non-artists now have the ability to play around and create outstanding pieces just using text. What the A.I. engine does is to assemble a completely new image based on a massive database of images. The A.I. learns about these images, stores similar images together using networks of meanings and semantics. Despite how realistic the images you see here, these are not real scenes or objects. It contains no object value, the subject has no story, composition accidental. It is up to the human user to determine (and understand) what he’s looking at. Therefore, despite having A.I. do the hard work, it still needs human to feed it information and understand the output it regurgitates. This is where the Recursive Design Studio resides: Training the human

This semester, RDS will utilise MidJourney A.I. software. Subscription is USD30 per month (shared per studio, for 2 months) Alternate plan is to use Stable Diffusion A.I., which is free but requires extensive hardware (depends on students’ equipments).