Space, Time, and Artificial Intelligence
Taubman College of Architecture + Urban Planning
Faculty member: Matias del Campo (Architecture)
Can artificial intelligence (AI) dream new visions of architecture into being? That’s the fundamental question of Space, Time, and Artificial Intelligence. Students in the class create AI neural networks that emulate mental processes such as dreaming and hallucinating, and then provide these AI computer programs with large sets of images from UMMA’s digital collection database. The AI is trained to understand the visual qualities of these images and then use those qualities as the basis for imagining new architectural structures and styles. The resulting designs are neither completely authored by a specific artist/architect nor by the algorithm, questioning agency and authorship in art and architecture.
Can artificial intelligence dream new visions of architecture into being?
We’re Letting a Neural Network Loose In an Art Museum’s Collections Database
Students in Professor Matias del Campo’s Space, Time, and Artificial Intelligence class are asking questions about the relationship between artificial intelligence, architecture, and creativity by creating an artificial neural network that will learn the various features of images in UMMA’s collection. Using that data, the neural network will then produce it’s own architecture structures and styles.Read More
Works included in this collection
In this course, the entirety of the UMMA collection will serve as data points for the neural network being developed. As the work progresses, we’ll use this space to share some of the machine’s work.
Lead support for this exhibition is provided by the University of Michigan Office of the Provost, Erica Gervais Pappendick and Ted Pappendick, the Eleanor Noyes Crumpacker Endowment Fund, and P.J. and Julie Solit.