MFA Thesis, Web Design, UX Research
The goal is to educate younger generations by conducting a systematic analysis of art movements, and then applying their characteristics to the AI-generated patterns.
Tools
AI Text-to-Image Generators:
Midjourney, DALL·E 2,
DeepAI, Stable Diffusion
Figma, Webflow, p5.js
Illustrator, InDesign
Overview
The use of AI technology is now prevalent in various fields, and it has become ubiquitous. Graphic designers need to stay up-to-date on new trends to deliver stories through compelling patterns. Beyond Yesterday Toward Tomorrow can aid users to have a stronger understanding of the features of patterns and the art movements they represent.
Problem
AI generated patterns may not be an accurate representation of art movements.
Patterns play a significant role as visual elements. They provide the foundation for a meaningful, cohesive, and well-balanced work of graphic design.
Starting Point
My initial inspiration came from an article discussing the "Nostalgia Pendulum," a concept suggesting that popular culture cycles every 30 years. At first, I agreed with the 30-year cycle idea, but then I questioned whether young generations understood the nostalgic feelings. Plus, as AI technologies are trending, I wanted to find out how AI technology would affect graphic designers, specifically in creating patterns which originated from the past.
Exercise 01:
Comparing AI Generators
In this project, I embarked on exploring four different AI text-to-image generators: Midjourney, DALL·E 2, DeepAI, and Stable Diffusion.
Comparing four different AI text-to-image generators provided useful insights into how AI generates images based on the keywords users input. After studying the patterns, I realized that they didn't show key features of art movements.
Exercise 02:
Drawing Outlines of Shapes on Images
In the second exercise, participants were asked to draw outlines on AI images after I introduced my thesis about exploring the relationship between AI technology and past visual language in patterns. I assessed participants' ability to recognize art movement features and their familiarity with art movements.
The results obtained from the exercise were very diverse, indicating that each participant had a unique interpretation and perspective, even though the images used were the same.
Exercise 03:
User Testing through App
I conducted a user testing with a mid-fi prototype of the app with graphic design students. Through this test, I aimed to assess their understanding of the true meaning of art movements, as well as, to determine whether they enjoyed and learned from the interactive activities.
The first user testing was to observe how participants extracted shapes from AI generated images, which were based on the keywords and art movement users chose. Afterward, they could enjoy a one of a kind filter composed with the shape they created and learn about the meaning of the art movement and artists.
The second user testing, a reverse version of the first test, was to observe how participants chose the shapes from the filters created by others. By scanning the QR code with a camera, users can explore the extracted shape and its AI generated origin pattern. They can get further information about the art movement.
The common feedback was that participants could understand the true meaning of the art movement. However, they found it text-heavy and challenging to absorb all the context. As a result, I realized that creating a website, rather than an app, would be more suitable. In this case, I could separate the interactive activity to different sections on the website. Additionally, I should incorporate illustrations to better convey the key points.
Website
This experience guides people to explore AI patterns by picking an art movement. Through the website, users can learn to recognize key pattern features and assess how relevant AI generated patterns are to their origins.
↘ Explore the website :
BEYOND YESTERDAY TOWARD TOMORROWSolution
Bridging the Gap: Empowering Artistic Awareness through a Comparative Exploration of Art Movement Origins and AI Images on an Interactive Website
It is important for individuals to be mindful of the origins and unique characteristics of different art movements when using technology to create or interpret images.
Activity
This website features an activity section where AI is employed to enhance the interactive experience. The target audience encompasses not only graphic designers, but also regular individuals.
This application utilizes extracted shapes and then forms an image of whatever the camera detects. Through the activity, I hope people will become more interested in graphic elements and design. It also has potential to educate people and generate interest in the app and website.
Shapes
The pattern used in the activity was generated from an AI image using keywords derived from art movements. Shapes extracted by people were randomly combined to create a new pattern.
Conclusion
"Beyond Yesterday Toward Tomorrow" can assist users in developing a strong understanding of pattern features and the art movements they represent. By conducting a systematic analysis of art movements and applying their distinctive characteristics to the AI-generated patterns, graphic designers have the potential to revolutionize their designs. There are unlimited possibilities when graphic designers learn how to fully take advantage of AI technologies to enhance their work.
Main Challenge
AI text-to-image generators are trending, and my intention was to understand how they interpret patterns using art movement keywords. It required a more objective perspective, making it difficult to define and analyze the style of art movements accurately. Nevertheless, I was able to read books and articles to compare the original patterns from art movements with those generated by AI text-to-image generators. Additionally, it was exciting to organize accurate information for graphic designers.
Next Step
There are opportunities to expand the types of art movements covered. Additionally, it could include more interactive and educational information.