Project Description:
Opening the Black Box: Making A Tactile Model of How AI Works
Professor Jiayi Young, Department of Design
Project timeline: (now - June 2027; first working prototype by October 2026)
This project is a tactile, physical model that makes the inner workings of contemporary AI systems—specifically, U.S.-developed large language models—legible to the public. The preliminary thinking (open to suggestions) is that the model will be structured around four core processes: data ingestion, model training, probabilistic reasoning, and output generation. Sensors and Microcontrollers can be embedded in the model as appropriate. The goal of the project would be to translate abstract computational operations into material, tactile, and sensory form. This model will serve first as an educational tool and also provide opportunities to ask critical questions about the impact of AI development on society.
A Spatial Interface for Self-Realization
Kyle Taniguchi (undergraduate Design major under Professor Jiayi Young supervision)
Project timeline: (now - October 2026)
This project is an interactive projection-mapping installation designed to help individuals understand and break addictive habits by reflecting on personal triggers. Triggering elements (e.g., people, locations, objects) are projected onto physical blocks in space. Participants place and move these blocks. Their movements are tracked and used to simulate corresponding real-life scenarios, which are projected onto the blocks and surrounding floor. As the blocks change position, the system generates new projections in real time. The design is based on the clinical RAIN (Recognize, Accept, Investigate, Note) mindfulness framework proposed by Judson Brewer. It also incorporates principles from Just-in-Time Adaptive Interventions (JITAIs) to support self-awareness and behavior change.
Technically, the system would integrate generative AI to create visual scenarios based on participant input and to generate variations that support reinforcement learning. The project is seeking engineering collaboration with an ECE student to program the camera recognition system and real-time genAI graphics.
A related example of spatial interaction can be seen in Giant Connecting Block Town by teamLab https://www.teamlab.art/w/giant-block-town/.