Imagining AI native map canvas that turn fragmented searches into seamless journeys

Timeline

Timeline

May 2025 - Aug 2025

Company

Company

Google Maps helps billions of people navigate, explore, and understand the world through accurate, real-time, and visually rich map experiences.

Role & Team

Role & Team

UX designer collaborating closely with two other designers, an adjacent design team, PMs, as well as engineering stakeholders.

Responsibility

Responsibility

Led the exploration of how the map canvas and its elements could dynamically adapt to AI-powered queries — from route-specific to place-specific requests — ensuring the information shown was timely, relevant, and helpful. Built a generative UI framework that served as both guideline and visual prompt, ensuring AI-generated maps communicated information effectively while following core UX principles of clarity and usefulness.

Context

Maps has long been the place you go to find "the nearest café" or plan "the fastest route home". Clear, powerful, dependable.

Maps has long been the place you go to find "the nearest café" or plan "the fastest route home". Clear, powerful, dependable.

Maps has long been the place you go to find "the nearest café" or plan "the fastest route home". Clear, powerful, dependable.

But the world’s questions are not always so simple.

But the world’s questions are not always so simple.

But the world’s questions are not always so simple.

For queries like these, language models speak in paragraphs, yet miss the map. Maps show the world, yet miss the story.

For queries like these, language models speak in paragraphs, yet miss the map. Maps show the world, yet miss the story.

For queries like these, language models speak in paragraphs, yet miss the map. Maps show the world, yet miss the story.

Screenshot of ChatGPT, Perplexity, and Google Map (Aug 2025)

The problem

so we asked, what if the map itself could the answer?

so we asked, what if the map itself could the answer?

so we asked, what if the map itself could the answer?

How might we leverage our AI capabilities to use map as a canvas to communicate complex answers more helpfully?

How might we leverage our AI capabilities to use map as a canvas to communicate complex answers more helpfully?

How might we leverage our AI capabilities to use map as a canvas to communicate complex answers more helpfully?

per company policy, project details cannot be shared publicly yet

please contact Eric at ericwanguxdesign@gmail.com
for more information

for googlers, feel free to check out go/ericwang for my work