After reading the book Tiny Experiments by Anne-Laure Le Cunff, I started noticing patterns in my design interests — skills I wanted to sharpen and questions that kept resurfacing. I’ve always believed in learning by doing. So with my growing interest in how AI can support design, I framed this hypothesis:
Using AI to explore design areas where I want to deepen my understanding could accelerate my discovery of the products and industries I’d most like to contribute to.
This idea energized me. I quickly drafted a PACT goal: over one week, I’ll explore five tech products I’m curious about and design an improvement for each — learning by making, reflecting, and sharing.
Why Tiny Experiments?
For years, I’ve worked on complex, specialized products — often deep in Android’s System UI, novel interaction models for Assistive Technology, and accessibility of design systems used across hundreds of Google products. Recently, though, I’ve been craving something different: the joy of learning for its own sake. The freedom to ask small, open-ended questions. To explore without over-polishing. To invite others into the process.
I found these tiny experiments helped me:
- Reframe design as a playful practice, not just a professional output
- Learn faster by testing, sketching, and sharing in shorter cycles
- Collaborate with AI tools in more creative and intentional ways
What I Explored
Each experiment centered on a design theme or question that’s been tugging at me. Some came from personal friction points. Others were sparked by curiosity or a desire to revisit areas I’ve loved but drifted from. Topics include:
- Semiotics: How symbols and metaphors shape meaning
- AI interaction patterns: How GenAI might be better integrated into everyday tools
- 3D and drawing tools: New workflows for spatial and 3D model-based design
- Growth and onboarding: How emerging products handle engagement and user ramp-up
Rather than focusing on one domain, I let my curiosity guide the scope, and took notes of what I learned.
Tools I Used
I explored with a range of tools and fidelities, each influencing my thinking in different ways. Some were fast and messy; others more structured. I used each tool with a specific intention:
- ChatGPT: Conversational ideation, voice transcription, narrative shaping
- MidJourney (Draft + Conversational Mode): Rapid visual design
- Firebase Studio: Low-fidelity functional prototyping
- Mind Maps: Still undefeated for making sense of messy ideas
What to Expect from This Series
Each article in this series captures a single, small experiment. Including:
- The product or design challenge I tackled
- A moment of friction or curiosity that sparked the idea
- A breakdown of what I tried (and what didn’t work)
- Notes on how AI tools helped or got in the way
- Visuals ranging from sketches to prototypes to AI-generated outputs
- A few takeaways or questions I’m still holding
These aren’t case studies — they’re working notes. I’m sharing them to learn in public, connect with others exploring similar territory, and build a habit of writing along the way.
If you’re experimenting with GenAI in your creative or design process, I’d love to hear what you’re uncovering too.
Let’s dive in.
More in this series
- My First Tiny Experiment: Onboarding that Builds Momentum
- Can AI Teach Design Theory? Designing Strain and Recovery Icons for my Fitness Wearable
- Coming soon…
I’ll update this list as I publish each article.