Authors
Matias del Campo, Sandra Manninger, Leete Jane Wang, Marianne Sanche
Keywords
Artificial Intelligence, Agency, Style Transfer.
Abstract
The main aim of this paper is to demonstrate a methodology of design that interrogates aspects of Artificial Intelligence and its abilities to develop novel sen-sibilities. The presented project, The Church of AI (fig1), discusses the design technique itself as well as the underlying aspects of aesthetic, ethic, and existence. The project started in the course Architectural Automations. An Advanced Design Studio of PennDesign1, with the focus on the transforming potential of Artificial Intelligence and automation in architecture. The results of the studio addressed aspects of autonomous behavior in construction (Performative Machine2), the changed relationship to creative practice (Golden Playhouse3), and Worship & AI (Church of AI4). The later was selected as a topic for this paper as it demonstrates in a provocative fashion the multitude of lenses of observation for a problem like Artificial Intelligence and Architecture. Not only as a toolset to optimize very spe-cific elements of architecture such as floorplan, material consumption, and struc-ture, but rather to emphasize architecture’s ability to serve as a cultural marker and place of worship. In that extent it proposes a position that radically challenges the idea of computational methodologies as tools of expediency and efficiency and ra-ther embraces the possibility to use it as a tool of communication between the human mindset and an, as to this date, alien intelligence. Alien in the sense of defamiliarization or estrangement5. Following intense conversations in the studio about the nature of AI, and its possible impact on architecture the project The Church of AI started speculating about the various possible aesthetic conversations possible through the use of AI. Of course, AI is a generalist term that includes a wide range of approaches and ideas. In order to make progress within the frame of one semester a more specific approach was needed. The main question was whether an AI could create a novel sensibility based on specific datasets, and how human intervention could steer the results. Exploring possibilities for a Human/AI collab-oration.
