MFA Graduate Research Project | Lukas Emory, You Eye

Lukas Emory is a full-time Creative Manager in the EdTech industry who enjoys teaching web and interface design to college students. Outside of work, he spends his time reading, writing, and exploring ideas that connect technology and creativity. The KUCDMFA program sparked his curiosity about artificial intelligence and its growing role in the world of art and design. Lukas believes good design doesn’t need to change the world—as long as it’s fun to make or fun to use, it’s doing something right.

You Eye is an interactive web application and research project created by Lukas Emory as part of his graduate thesis exploring the role of artificial intelligence in art and design education. The project investigates how AI can be introduced to creative classrooms not as a replacement for human decision-making, but as a tool to enhance critique, accessibility, and reflection in the design process.

Inspired by discussions around the calculator’s once-controversial introduction into math education, You Eye uses both human-trained algorithms and deep-learning models to provide real-time feedback on user interface designs. Emory trained the program to assess font size, color contrast, and design clarity using traditional design principles and accessibility standards, then layered in OpenAI’s GPT-4 to provide creative critiques based on both the image and user-submitted goals.

Users upload a screenshot of a mobile or desktop interface and receive an immediate, four-part analysis: a heatmap indicating too-small fonts, warnings about low contrast text, a tone description of the design’s color palette, and a one-sentence AI-generated classification of the interface as functional, informative, or decorative. The end result is a design critique that feels fast, focused, and rooted in both code and creativity.

The goal of You Eye was to reframe AI as a partner in the classroom—one that can support feedback, self-reflection, and the iterative process of design without taking the human out of the equation. The live web tool is fully functional and was built with Flask, OpenCV, EasyOCR, and OpenAI’s API.