Research

Eskenazi Technology Innovation Lab Research Projects

AI's Impact on Human Creativity Through Origami Design

Jiangmei Wu (Interior Design), Garim Lee (Merchandising), Ran Huang (Merchandising)

In a collaborative study bridging design and social sciences, a team led by Professor Jiangmei Wu investigates the intersection of artificial intelligence and human creativity by creating a Generative Origami AI and study students’ use in an advanced architecture drawing class. The project, which leverages open-source and Diffusion Models (DMs), aims to address the ongoing debate about AI's potential to augment human creativity.

The study focuses on how AI tools can enhance the origami-inspired design created by college-level students. By comparing traditional design methods with AI-assisted approaches, the researchers hope to gain insights into how generative AI technologies support the creative process in areas that require an understanding of material properties and tactile exploration. The project utilizes established creativity assessment tools and examines factors such as prompt linguistic style and perceived tactility to provide a comprehensive understanding of AI's impact in real-world design scenarios. Currently, the team is developing the Generative Origami AI, a web-based app utilizing the NSF funded resources such as Jetstream2 that provides a library of virtual machines and shared software designed to allow researchers to perform discipline-specific research. With the Generative Origami AI, users can quickly explore customized design alternatives and make origami-inspired designs powered by generative AI.

Artist Unveils AI-Powered "StoryCatcher" Exploring Displacement Narratives

Megan Young (Digital Art)

By a fusion between artificial intelligence and participatory art, artist Megan Young has unveiled "Carry," an AI-powered "StoryCatcher" that brings personal narratives of displacement to life. This creative project, which combines Python-based natural language processing (NLP) and retrieval-augmented generation (RAG) techniques, draws from a growing repository of personal stories collected through one-on-one conversations with women from around the world. Carry operates as a dynamic, interactive archive within gallery and museum settings, engaging viewers in conversations about her journeys while inviting them to share their own experiences.

Young's project, supported by a team of data scientists and researchers through The Program for Faculty Assistance in Data Science (FADS), pushes the boundaries of empathy and connection in the age of AI. The project, which continues to evolve based on ongoing research and participant feedback, is being presented at various venues including the Grunwald Gallery, the University of North Carolina Wilmington CAB Gallery, and as a Cleveland public art project. By treating the AI as if it were raised on collective oral histories, Young creates a hybrid, technology-rich storytelling experience that challenges viewers to reconsider their understanding of displacement, shared experiences, and the potential of AI in preserving and sharing human narratives.

Who Designs Better? Crowdsourcing Jurors to Assess AI-Assisted Performance

Hoa Vo (Interior Design)

Professor Vo’s studies aim to shed light on the complex relationship between AI and human creativity in design processes. Her study, "Who Designs Better? A Race Between Human and AI or Happy Co-Design," has recently concluded its data analysis. This project analyzes and compares lighting products generated by 1) AI, 2) human designers using AI, and 3) traditional human designers. Using the Creative Product Semantic Scale (CPSS), the study revealed significant insights regarding the differences across the designs. 120 Amazon Mechanical Turk workers (experiment 1) and 126 Prolific workers (experiment 2) evaluated these designs as they progressed from 2D sketches to 3D renderings and finally to immersive Virtual Reality (VR) models, providing a comprehensive assessment of each design's creative potential.

Building on these insights, a second ongoing study called "Muse-Gen" is expanding the scope to examine AI's impact on large-scale, complex spatial designs. Professor Vo aims to develop a generative AI model capable of producing detailed and credible museum floor plans. The study will then evaluate the creativity of these AI-generated designs through a combination of crowd-sourced assessments and expert ratings. While still in progress, Muse-Gen is expected to make significant contributions to our understanding of AI's evolving role in the design profession.

Studies Reveal Complex Relationship Between AI, Design, and Consumer Attitudes

Garim Lee (Merchandising)

Professor Garim Lee’s recent research examined the factors influencing consumer skepticism and liking of AI-created content. Lee analyzed Reddit data and surveyed 281 U.S. participants, finding that perceived appropriateness and novelty affected both skepticism and liking. The study revealed that the public’s ambivalence toward AI-generated ads and AI artwork shows different dynamics. One of the papers was presented at the 2024 Global Fashion Management Conference, where it received the Best Conference Paper award. Lee's research suggests differentiated strategies for managing consumer perceptions of AI-generated content in commercial versus noncommercial contexts. For commercial AI content, such as customized ads or product demos, marketers should focus on addressing skepticism by emphasizing ethical practices and regulatory compliance. For noncommercial AI content (e.g., artwork or museum displays), the priority should be increasing appreciation by highlighting innovation, uniqueness, and human effort in the creation process. These targeted approaches can help marketers effectively manage consumer perceptions of AI-generated content in various settings.

Generative Design for CNC Mold Machining in Slip Cast Ceramic Production

Justin Bailey (Creative Core)

Professor Justin Bailey's project seeks to advance ceramic slip cast production by incorporating CNC milling and generative design into the mold-making process. Unlike traditional slip casting, which relies on plaster molds created from positive forms, Bailey’s innovative method uses 3-axis CNC machining to directly mill molds, eliminating the need for those forms. This approach helps avoid undercutting issues, facilitates the integration of mold elements, and offers greater design flexibility. By utilizing Grasshopper for Rhino3D, he is creating parametric designs that produce both the ceramic objects and their molds as 3D models, resulting in complex multi-part molds.

Bailey's goal is to establish a system for producing series of molds that enhances efficiency in creating modular ceramic works, expanding the production capabilities of slip casting and allowing for greater output across various applications. Ceramics are versatile materials that provide diverse glazing options and durability, making them suitable for food-safe, water-safe, and architectural uses. By merging traditional techniques with digital fabrication and generative design, Bailey's research aims to bridge artisanal craftsmanship with industrial production, potentially transforming the design and manufacture of ceramic objects for both aesthetic and functional purposes.