rene | SCIO

Returning to the topic of generative visuals: the property of Ai to generate hipper realistic imagery is maybe more popular, but what about accidents and inaccuracies? Can the AI produce useful imperfections, much like an artist who uses the accident as a tool. The comparison of the AI imagery to the Art object is misplaced, of-course, however such parallel considerations can help us explore the creative process. The argument isn't if AI can produce art. The question is what is it that AI can produce that is different, useful that isn't related to pure automation.

The series of images presented in this small experiment are a selection of exactly such accidents. Floating objects, disoriented compositions produce almost pre-Albertian visual systems, before the application of perspectives and the strong hierarchy of space and objects within the image.

Most curiously, the process is not exactly controlled by the author. How much control or decision is involved in the making of the design object? When and how those decisions are to be made? Is design at all a domain of decision-making, or rather a gradual narrowing down of the possibilities that the context allows: a concretization of collected internal and external factors...

//prompt < an image of an inside of a flying castle, featuring its vast interior that looks like a flying balloon, renaissance fresco panted on the walls, birds flying and parts of Palladio's theater >

Published by scio / 2023
Topic: architectural design/ Ph.D research project

Institution: ENSA Nantes
supervisor: L. Lescop
Lab: Crenau

*all images are generated by Ai