The project explores and challenges the recent boom of image-generating Al technology while speculating on the transformation of derelict (decaying) buildings found throughout the site into entirely new aggregations and programs. Tackling the overall complexity and unpredictability of initial material deposits as a source for new emerging structures via the same complex and (un)predictable behaviour of Al-based text-to-image algorithms.
Every object that is created by composing non-uniform residues into a structurally stable aggregation is therefore different and goes beyond the obedience of the context. Yet, slightly expressed traces of the building it consists of are still present—referring to the surroundings and rejecting them at the same time, eventually allowing pedestrians to interact with the same disjointed "materiality" of the familiar context they are used to pass by.
AI Preliminary Sketches
letting AI dream on top of the sketchy defined physical model, using Stable Diffusion 1.5 (ControlNet) and GEN-1 (Runway)
AI Concept
Masterplan Design Concept
The irregularity in the placement of the structures came from manipulating the distinctive vernacular grid of Viennese landscape designs of the Baroque period. Post-processing the dataset of existing parks through Al-based Neural Networks (GAN) resulted in a new morphogenesis of compositional axes superimposed onto the urban layout of the site.
Presented at the "Wien>>AI" Exhibition in Vienna, Austria
curated by Hani Rashid
Finally, the new grid laid over the shattered and stagnant area of Heilignestadt produces a new patterned landscape of follies that joins the formerly separated east and west part of the site into a new whole
Published at the DEZEEN, School Shows 2023 - University of Applied Arts Vienna presents 15 architecture student projects
Resin 3d Printing WIP
SS2023, Studio Hani Rashid, Die Angewandte
