More Than Human explores how deep learning can aid in developing novel design tools. It continues the biospatial design research agenda of Research Cluster 7, challenging the paradigm of building first and landscape second. It rejects the long-standing conceptual separation between humans and the natural environment and aims to reintegrate non-human agency into architecture.

Extensive experimentation with procedural dataset design and generative neural networks has culminated in the creation of a sketch tool that integrates both human and non-human spaces as lines are drawn. Simple human input is interpreted by a neural network and processed through a series of site-specific environmental analyses, producing a massing model with embedded ecological intelligence. This model represents a 3D map of the characteristics of the site and can be used to inform developed architectural designs.