By The Metropolitan Museum of Art
Brett Renfer
Evolving over time
Over our 156 year history, The Met has witnessed—and been a part of—many radical shifts in the way visitors experience art.
Exploring parallel futures
Each technological leap offers many paths forward, often offering a huge array of possibilities, ideas, and challenges.
Generative AI is one such leap, and, since it's a very broad and transformative technology, the number and scale of possibilities is dizzying.
Our idea: explore this complexity head-on through a new, focused position, a "Technologist in Residence." We decided to start with the on-site experience, and set out to develop as many experiments as possible: what’s working? What’s not? What do our visitors want?
User-Testing the Unlikely Pairings Prototype by The Metropolitan Museum of ArtThe Metropolitan Museum of Art
6 Months, 3 Sprints, 20+ Prototypes
We embedded a creative technologist within the Met for six months to explore how AI might impact the visitor experience at the museum. They rapidly built prototypes using Gemini and Vertex AI, and tested them on the floor with visitors, exploring both what works and what our visitors want and need.
The sprints
We broke up the 6 months into 3 focus areas:
"Wayfinding," helping visitors look broadly across the museum to find what's relevant and of interest
"Close Looking," diving deep into an individual work or gallery
"The Met and the World," connecting to data and experiences outside our walls.
GenAI supercharges iteration
AI coding agents made it feasible to workshop a concept Monday, build a working prototype Tuesday/Wednesday, test on Thursday, and document on Friday.
This, on top of a defined process and support structure, made this pace and process possible.
Our experiments led to a broad array of insights. One above all stood out.
User-Testing the Wall Label Prototype by The Metropolitan Museum of ArtThe Metropolitan Museum of Art
Transformative AI wins the day
Our visitors were excited about projects that took full advantage of AI: offering content in dozens of languages, adding opportunities for deep personalization, etc. They were skeptical of ideas that don't meet that threshold.
User-Testing AI Prototypes by The Metropolitan Museum of ArtThe Metropolitan Museum of Art
Dive in
In the following three posts, our resident Julia Daser will dive deep into each sprint: what worked, what didn’t, and what we learned from our visitors. Enjoy.
You are all set!
Your first Culture Weekly will arrive this week.