Leon Honge
Research Mentor: Audrey Bennett
Mentor Department: Not Available, ART
Author(s): , Audrey Bennett, Ron Eglash, Leon Honge
Session: Session 3 (11:00 AM – 11:50 AM)
Presentation Type: Poster 25
Abstract
With the rise of Artificial Intelligence, class bias is becoming evermore prevalent in society. AI is designed as a “helper†for lawyers, engineers, and other professionals, but presents itself as a job threat for low income labor. Our research team takes the opposite approach, by working with small businesses in Detroit to identify AI systems they would like to see implemented. Just as importantly, we implement the AI on a worker-owned platform, Artisanal Futures (artisanalfutures.org), so that profits go back to the grassroots community, not corporations. My research began with the observation that many worker-owned businesses in Detroit are using upcycling, a method used to transform waste or old scraps to create new products. Thus, in an effort to empower these small businesses, an AI image generation tool was created, using upcycling principles to transform their original product ideas into virtual designs. Within this tool, we aimed to design an intuitive GUI to ensure that these artisans could easily access the full resources of our AI model. The initial GUI served as a provocation exploring how this application could fit into their craft without damaging creative enjoyment. Technically, the AI model offers a variety of available parameters and controls, but that is overwhelming for any lay user. Throughout multiple feedback rounds and redesigns, we consolidated the layout into different utilization “modes†with specific, comprehensible vocabulary (Quick Try, Structured, Pre-Trained), understood the needed workflows, (e.g. model training for subsequent image generations), and added features such as model weight sharing for community collaboration between artisans. By remaining responsive to user needs, this iterative GUI design seeks to become a technological means of empowerment of these smaller businesses. Just as commercial AI companies focus on empowering scientists, lawyers, and doctors, our application focuses on empowering these small businesses to compete and thrive.
