Skip to main
University-wide Navigation

In his lecture Building From Source, Patrick Danahy will explore the potential of AI-driven feature embedding as a tool for designers to challenge and expand their perceptual frameworks. Danahy will also lead a 2-day workshop in which participants will learn how to input various design prompts and adjust parameters to generate unique design elements using AI-driven diffusion models. 



lecture graphic

Building from Source: AI and Feature Embedding

Lecture: Friday, April 11 @ 4:00 p.m. | David Biagi Forum

The world around us is complex, constantly evolving in ways that often go unnoticed. Our traditional understanding of materials, forms, and spaces tends to be based on surface-level attributes—what they look like, feel like, or how they function in a specific context. This approach, shaped by ingrained assumptions, can limit our ability to perceive the deeper, interconnected systems that influence the dynamics of our environments. As a result, designers often overlook the invisible relationships that govern the behavior of materials and spaces.

This lecture explores the potential of AI-driven feature embedding as a tool for designers to challenge and expand their perceptual frameworks. By leveraging artificial intelligence, particularly machine learning, designers can go beyond the obvious and uncover hidden, high-dimensional features within objects and environments. These AI techniques allow for the discovery of complex relational systems and networks that were previously inaccessible, offering new insights into how materials behave, interact, and influence larger systems.

It will explore how AI algorithms can analyze seemingly simple materials to reveal their microscopic properties, behaviors, and relationships within broader contexts. This deeper understanding enables designers to move beyond preconceived notions and make more informed design decisions. The lecture will also demonstrate how AI-driven design can identify patterns in data to predict how materials interact, and how spaces function. These insights can lead to dynamic designs, grounded in a deeper understanding of underlying environments.

Ultimately, it will demonstrate how embracing AI allows designers to uncover new relationships, create more responsive designs, and engage more fully with the complexity of the world around us.



workshop graphic

Machined Models: Advanced AI in Architecture

Workshop: Saturday, April 12 and Sunday, April 13 | 10:00 a.m. - 6:00 p.m. | Location TBD

In the age of AI-driven design, new tools like diffusion models are revolutionizing the creative process, allowing designers to generate complex visuals and innovative forms from simple inputs. One such platform that has gained popularity is Comfy UI, which leverages the power of diffusion models to provide an intuitive interface for AI-assisted design workflows. In this workshop, we will explore how to harness the capabilities of Comfy UI for design applications, from concept generation to refined outputs.

Through the use of Comfy UI, participants will learn how to input various design prompts and adjust parameters to generate unique design elements using AI-driven diffusion models. We will walk through the steps of setting up your workspace, generating preliminary designs, and refining them into more specific, usable assets for projects. In addition to the core design process, the workshop will explore integrating AI-generated content with traditional design tools, allowing participants to blend computational power with human creativity seamlessly.

We will delve into how diffusion models can be used for architectural visualization, material design, and even generating concept art, all while adjusting model parameters to fit different aesthetic styles or functional requirements. By the end of the workshop, participants will have hands-on experience with both the creative and technical sides of using AI models in design workflows, from generating concepts to preparing assets for final production.

Prerequisite for this workshop is a basic understanding of design software and workflows. Familiarity with AI tools or graphic design software will also be helpful but is not required. Participants are encouraged to bring their own design ideas, as the workshop will include time for hands-on exploration and experimentation within Comfy UI.