Role
Founding Product Designer
Tools/Skills
Figma, Prototyping, iOS
Team
1 Software Engineer
Timeline
March 2025 -
I’ve been conducting research with Professor Chilton and the team on training AI to understand and generate humor — gathering and scraping comedic data to study how models interpret timing, absurdity, and tone. These findings shaped how we approached the Humor Project’s design and framing around “what makes AI funny.”
Through this research, I wanted to design an experience that makes AI humor feel fun to consume, not just generate — blending data-driven insight with the social mechanics of entertainment apps.
AI IS JUST NOT FUNNY
Despite its rapid progress, large language models still struggle with humor’s emotional nuance and originality.
CONTEXT
Understanding how humor travels across platforms.
RESEARCH AND INSPIRATION
STUDYING SOCIAL HUMOR FLOWS
Reddit’s nested humor threads informed comment interactions
TikTok’s quick feedback loop inspired the ‘react instantly’ flow.
Tiktok
DESIGNING A FUNNY, SOCIAL AI EXPERIENCE
Combine generation and consumption — users can both browse and create.
Encourage engagement loops through reactions, saves, and sharing.
Keep the humor curated, relatable, and culturally aware.
As the founding designer, I aimed to build an iOS app that lets users browse, react to, and share AI humor effortlessly, while also creating content within the same flow.
GOALS
QUESTION
These six insights helped us identify what makes humor feel human, social, and worth sharing — and shaped the foundation of the product.
NEEDS
WHAT WE LEARNED ABOUT HUMOR
Humor rewards creativity
Snackable content travels further
Tap, scroll, react → no cognitive load
Collaborative humor increases engagement
Not everyone laughs at the same thing
Humor spreads fast in short formats
Humor needs low effort to consume
Humor feels fun when co-created
Humor is subjective
Users want to remix, caption, or add their spin
Humor is social currency
People share jokes to connect + signal personality
Make humor feel lightweight, replayable, and worth sharing.
CONCEPT DEVELOPMENT
DEFINING A NORTH STAR FOR AI HUMOR
DESIGNING A FUNNY, SOCIAL AI EXPERIENCE
Combine generation and consumption — users can both browse and create.
Encourage engagement loops through reactions, saves, and sharing.
Keep the humor curated, relatable, and culturally aware.
As the founding designer, I aimed to build an iOS app that lets users browse, react to, and share AI humor effortlessly, while also creating content within the same flow.
GOALS
QUESTION
MAKING HUMOR INTERACTIVE
MAKING HUMOR INTERACTIVE
As the founding designer, I aimed to build an iOS app that lets users browse, react to, and share AI humor effortlessly, while also creating content within the same flow.
DESIGN ITERATION
Designed and prototyped an iOS humor app with full interaction flows.
Created a scalable framework for humor curation and user reactions.
Established a design system for tone, timing, and caption placement.
This early direction focused on clarity and hierarchy—creating a lighter, faster interface that still supported tagging, labeling, and note-taking without overwhelming the user.
Continue model fine-tuning with Professor Chilton’s team.
Onboarding
@ Frich
Next Steps
Launch closed beta for testing humor quality and social behavior.
Outcomes
Explore expansion into short-form AI sketches or meme remixing tools.
OUTCOMES AND NEXT STEPS
DID THE JOKE LAND? + WHAT’S NEXT