BACK

Live Website

OVERVIEW

HIGHLIGHT

PROCESS

PROCESS

TAKEAWAYS

Young Ones 2026 MERIT

Generative AI

Creative Concept, Motion

Gemini: Because learning isn't One Size Fits All

Gemini: Because learning isn't One Size Fits All

OVERVIEW

This project explores a new creative concept for Google NotebookLM, showing how Gemini can help students turn teacher-provided materials into personalized study tools. By solving one student’s learning challenge, the concept illustrates how adaptable learning can create broader impact across education.

This project explores a new creative concept for Google NotebookLM, showing how Gemini can help students turn teacher-provided materials into personalized study tools. By solving one student’s learning challenge, the concept illustrates how adaptable learning can create broader impact across education.

TIMELINE

3 months

ROLE

UI motion & Strategy/

Research

MY TEAM

Yuriko Zheng

Joyce Ahn

TOOLS

Figma

After Effects

Midjourney

TIMELINE

3 months

MY TEAM

Yuriko Zheng

Joyce Ahn

ROLE

UI motion & Strategy/

Research

TOOLS

Figma

After Effects

Midjourney

HIGHLIGHTS

AI GENERATION PROCESS

Creating Our Characters

Creating Our Characters

Gemini was used to generate the initial static characters.


We refined these visuals in Photoshop to create stronger references, which were then used as prompts in Midjourney and Reve to generate motion.


Finally, Adobe After Effects was used to bring all elements together and assemble the final animated sequences.

TAKEAWAYS

Design is not a one way street.

VRBO solves finding a place, but not planning the trip.

Prompting is a design skill.

Prompting is a design skill.

Vague inputs produce vague outputs. The more intentional and specific the prompt, the stronger and more usable the result: the same principle that makes good design briefs work.

Vague inputs produce vague outputs. The more intentional and specific the prompt, the stronger and more usable the result: the same principle that makes good design briefs work.

The best tools disappear into the work.

The best tools disappear into the work.

Moving across AI generation, refinement, and motion design taught me to connect different tools into one cohesive outcome. What matters isn't the stack, it's whether the result holds together.

Moving across AI generation, refinement, and motion design taught me to connect different tools into one cohesive outcome. What matters isn't the stack, it's whether the result holds together.

Not every student learns the same — and design should say so louder.

Not every student learns the same — and design should say so louder.

Using generative AI to build the visual direction pushed me to let color, composition, and imagery carry the message: that one tool can meet students wherever they are. When the words step back, every visual has to say it on its own.

Using generative AI to build the visual direction pushed me to let color, composition, and imagery carry the message: that one tool can meet students wherever they are. When the words step back, every visual has to say it on its own.

What's Next?

What's Next?

Reimagining NotebookLM’s art direction helped me explore how storytelling and simple visuals can make AI tools feel more approachable for students.


Next, I’d expand the concept to different subjects and specific learning situations to see how personalized learning could scale across classrooms.


AI should support learning, not replace it—helping students engage with material in ways that work best for them.

Thanks for viewing! If you want the whole case study, email me at yuriko_zheng@fitnyc.edu :)