Wada Quilts
Designer | p5.js
LLM for Coding Assistance:
Chat GPT
For my MCAD MAGWD Experimental Interaction Course, I created a pattern generator using p5.js that returns quilt block patterns. Each quilt has three colors, based on the 3-color combination section of the Japanese Dictionary of Color Combinations, Volume 2, by Sanzo Wada.
Scope
Train and provide design direction to an LLM to create a single-point user input interaction using p5.js that generates and exports graphic elements.
Deliverables
At least 16 cohesive graphic element outputs (images, audio files, text strings, stories, etc.) created by a collaboration between myself as designer and my trained AI system: Google Gemini. Documentation of my process.
Background & Inspiration
I’d been practicing color mixing, inspired by this cult favorite color dictionary. I wanted to use the outputs of this prompt to generate new patterns, inspired by classic quilt patterns.
My mother, grandmothers, and several of my friends are quilters. This project gave me a way to participate in creative quilting circles, just in another medium.
Challenge: Pattern recognition
Anything that wasn’t a very simple geometric pattern was hard for both Gemini and ChatGPT to draw. This shows my attempts to prompt the LLMs, using both verbal prompts and image inputs.
Note: These outputs are from summer 2025.
