My AI Teaching Assistant

My AI Teaching Assistant

Introduction

Thank you for having me here! This is a really incredible and interesting group to learn about, and I hope to participate in the future. My name is Dave Elving, an Adjunct Professor at the California College of Art here in San Francisco. Before that, I used to work as an interaction designer and creative director on Apple's Interac marketing team. I spent over a decade helping to build product websites, the Apple online store, the Apple Store app, and interactive experiences inside retail stores. I left that role in 2019 to pursue a passion for teaching at the California College of Arts.

In this talk, I want to share what I've learned about teaching with AI. I've kept this presentation short and, in light of what we've heard from Burton and Daryl, I have some positive, hopeful, and exciting ideas for designers using AI-enhanced tools. If there's time, I have a backup presentation, or we can just open up for conversation.

The Role of AI in Design

The image above was created using ChatGPT this morning. I simply prompted it with "show me an image of an AI teaching assistant in a college interaction design classroom." This was the first result, featuring a humanoid robot, an old-school blackboard, and an apple on the desk to signify a teacher.

AI is changing the landscape of design tools. One example is Figma, a tool becoming a standard in UX/UI prototyping for its collaborative capabilities. Figma recently announced an AI enhancement that allows users to generate a prototype frame almost instantaneously by inputting a prompt. While this is incredible, it often results in generic solutions that lack the refinement and thoughtfulness that human designers bring to the table.

Context and Taste in Design

AI lacks context and taste, elements crucial for effective design. Design isn't about copying patterns; it's about creating thoughtful, informed solutions to human problems. Therefore, the skills of design thinking, empathy, research, and context are more relevant than ever.

As Gruber pointed out with his "Figma is a ripoff engine" post, AI-generated designs can replicate existing patterns too closely, such as when Figma’s AI produced an app identical to Apple's weather app. However, this isn’t much different from how designers have always used templates. The skill lies in how we adapt these templates and materials to create something unique.

The Design Process

The design process is fundamental and universal. This process usually involves several stages: empathize, define, ideate, prototype, and test. AI tools like Figma can fit into this process, particularly in the prototyping phase, but human touch remains irreplaceable.

Principles for Future Design

Here are three principles I'm incorporating into my curriculum this semester, which I believe are relevant to everyone in the design field:

1. Start with People and Paper: Engage in direct interaction and use tangible materials like sticky notes for initial ideation and understanding.
2. Paper and Sketches Before Prompts: Before jumping into AI tools, it's essential to work out initial ideas through sketches, as they help refine thoughts and interactions.
3. One Demo is Not Enough: The ability to rapidly generate prototypes means higher expectations. Instead of stopping at one demo, aim to create multiple iterations for a well-rounded solution.

AI can be a powerful teaching assistant, but the core of teaching and design remains a human endeavor. These tools should empower us to do more and be better, not just faster, designers.

Keywords

• AI Teaching Assistant
• Design Thinking
• Figma
• Human-Centered Design
• Prototyping
• Interaction Design
• Context and Taste
• Empathy in Design
• Sketching and Paper Prototyping

FAQ

Q1: What is the main point of using AI in design according to Dave Elving? A: AI can aid in rapid prototyping and iterative design, but it lacks the context and taste that human designers provide.

Q2: How does Dave Elving suggest students should begin their design process? A: Students should start with people and paper—engaging with real users and sketching initial ideas on physical media.

Q3: Why does Dave believe one demo is not enough? A: With AI tools making prototyping faster, the expectation is to create multiple iterations to find the best design solution.

Q4: Why was the Figma AI tool criticized? A: It was criticized for generating designs that closely replicated existing apps, like Apple’s weather app, leading to accusations of it being a "ripoff engine."

Q5: How does Dave Elving relate traditional design practices with AI-powered tools? A: He mentions that designers have always used templates and prior examples; AI tools should be seen as an extension of these traditional practices to enhance creativity and productivity.