Interconnectivity of Deep Learning Models in AI-Driven Design Systems
Research Project and Publication — First Author, with co-author Daniel Bolojan.
Abstract: The work presented in this paper examines a new human-machine collaborative workflow, combining artificial intelligence (AI) — in particular learning systems' capabilities — and designer creativity in a comprehensive framework. In developing a new design workflow, we adopt systems theory and the need to break down the design process into its constituent components. That is, there is a need for multiple AI models connected to address the smaller (broken-down) design subtasks.
The research focuses on developing a design-system workflow, a "prototype" with interconnected AI and agent-based models (ABM) to address multiple architectural systems, at different design levels (design tasks, design phases), while providing the designer with varying degrees of agency. The paper showcases two case studies where this prototype is tested on two different design projects.