Date
Process Log 1 Focus
Embedded Files
8/27 - 9/3
Research Through Design: Nudging for Responsible AI
CONTEXT
AI features are heavily and gradually being embedded in most applications we are using. Interfaces are showing the "magic" possibilities of the AI feature on the screen asking users to use AI feature. Thesea features sometimes nudge users with or without data-driven information to trigger a behavior to use AI features. Can interface for AI features look different than magic wand or spark icon?
PROCESS
STEP 1. Finding WHY, WHAT, WHO, HOW.
In class, I explored why this research is needed, who are involved, what I can explore by making the prototype, and how I will learn through making the prototype.
STEP 2. I started searching icons for "AI" and "Generate" in image resrouce website like thenounproject or Freepik.
STEP 3. I made three different text boxes with different color and I searched different icons to represent "AI" or "Generate". I wanted to focus AI features specifically for designers at this stage.
STEP 4. Drawing Stakeholder map
I drew a stakeholder map just to see what people are involved in this topic. There would be 1. a company that uses LLM Gen AI model, 2. users, and 3.LLM GenAI model company that supports the model.
REFLECTIONS
I started wondering what kind of "Other" icons could be done just than "sparkle" or "magic wand" icon. What are the words that are related to help?
What icons (semiotics) think of "help", "generate", "seek", "create", "suggest" that explains AI functions?
The first keyword I used was "AI". Majority of the results were brain, data maps, orbit, chip, networks, and sparkle. The second keyword I used to search was for "AI icon" and the "sparkle" when I searched "AI icon", the results were more boring. Many icons represented the word document, file icons, and smile icons.
Then I used the word "Generate"- most of the icons were sparkles, and magic wands.
INSPIRATION/NEW QUESTIONS
By doing this experiment, I started to have more questions and tried to be reflective as possible. Being aware of all senses during the process led me to another questions that I had never done before. How could this be better designed? How did designers design UI before AI? What kind of intervention? and to whom? in where, what space?
I start to wonder why there are only sparkles and magic wand icons to express the word, "Generate"? Who was the very first one that made this icon? Why do they use gradation of purple and blue and animation to show it's process?
CONCRETE ACTION
I plan to research ethical principles of AI, responsible AI, transparency AI. I would like to know what ethical decisions can be made when people use AI generate button. Researching on what type of ethical dilemmas designers have will help me create more sharp prompts. I would like to find out how design application companies and Large Language Model AI companies share infomation on their AI principles.
Reference
Nolwenn Maudet, Anaëlle Beignon et Thomas Thibault, “Comment les entreprises de la tech nous forcent à utiliser l’ia”, 2025. Prototyping AI nudge interface for designers
https://limitesnumeriques.fr/travaux-productions/ai-forcing/en
The Proliferation and Problem of the ✨ Sparkles ✨ Icon https://www.nngroup.com/articles/ai-sparkles-icon-problem/
Thierry Charbonnel, “AI for designers”, 2025. https://www.figma.com/design/tE54EgaBtHr4xvANob2ihO/Ai?node-id=0-1
Copyright © Borami Kang 2025
Page updated
Google Sites
Report abuse