Explore The Met in new ways with generative AI

To mark 15 years of collaboration, The Metropolitan Museum of Art is unveiling two new generative AI initiatives with Google Arts & Culture.

For 15 years, The Metropolitan Museum of Art has collaborated with Google Arts & Culture to bring advanced technology directly into the galleries to transform how audiences connect with art. Guided by our shared belief that access to art should know no distance, we invited audiences to “virtually visit” The Met in 2011 and digitally explore the Museum’s Open Access collection in 2018. Now, we’re unveiling two generative AI initiatives designed to deepen how people discover art online and experience the museum in person.

We recently completed our Technologist in Residence program — a central focus of this milestone year. For this six-month initiative, The Met engaged creative technologist Julia Daser to investigate how generative AI can enrich each visit to the museum. Using Google Gemini and Vertex AI, Julia collaborated closely with museum curators to rapidly build and live-test prototypes directly in the exhibition spaces.

Over the course of the six-month digital residency, we built dozens of testable prototypes that address the entire visitor journey, from initial navigation to post-visit engagement. We’re excited about possibilities like improved wall text accessibility, wayfinding and deep post-visit engagement, and hope our findings can serve as inspiration for other institutions.

By Brett Renfer, Senior Program ManagerEmerging Technologies at The Met

2:56

In tandem with the residency, we are also launching Art Aura, an interactive, multimodal AI experience developed by the Google Arts & Culture Lab in collaboration with The Met. Powered by Google Gemini, Art Aura allows users to drag artworks, styles or descriptive phrases into a digital zone to uncover hidden thematic connections across the centuries of art featured in The Met’s collection, creating a personalized portrait of their unique artistic taste.

2:35

To discover more about these projects and continue your journey, visit the new Google Arts & Culture landing page for The Met and browse the Museum’s collection of over 200,000 digitized objects and 50 stories.

The Met: Prototypes & Play — Google Arts & Culture

You can explore The Metropolitan Museum of Art using cutting-edge generative AI through several interactive initiatives co-developed with leading technology labs. These projects transform passive viewing into an immersive, active dialogue with over 500,000 digitized artworks. [1, 2, 3, 4]

🎨 Discover Hidden Connections with Art Aura

Developed in collaboration with Google Arts & Culture Lab, Art Aura is a generative AI experiment powered by Google Gemini. [5, 6]

Interactive Discovery Zone: You can drag and drop text descriptions, specific styles, or images into a digital portal. [5, 7]
Cross-Century Mapping: The AI cross-references your inputs against hundreds of thousands of museum artifacts to map unexpected thematic links. [5, 8]
Personalized Art Persona: As you navigate the tool, it builds a shareable “Personalized Portrait” matching you to aesthetics like Cosmic Stage Designer or Classical Humanist. [5, 8]

💬 Chat with Historical Figures

The Met partnered with OpenAI to bring historical exhibitions to life using conversational language models. [2, 9]

Active Exhibition Narratives: During the Sleeping Beauties costume exhibition, a tailored version of ChatGPT allowed visitors to engage in live conversations with a 1920s wedding gown’s original wearer, socialite Natalie Potter. [2]
AI Photo Reanimation: Asking specific questions during the chat triggers an easter egg where the OpenAI Sora video model reanimates static wedding photographs into moving footage. [2, 9]

🔍 Next-Gen On-Site Museum Exploration

Through its Technologist in Residence program, the museum live-tested in-gallery prototypes built on Google Vertex AI. [10]

RAG Detail Analysis: Visitors can take photos of dense gallery objects and circle specific visual details. A Retrieval-Augmented Generation (RAG) framework instantly extracts and summarizes complex curator notes and essays into concise, easily readable insights. [11]
Smart Wayfinding: Generative models help clear the logistical hurdles of navigating the massive physical layout by synthesizing personalized museum routes based on real-time crowd pacing and individual interests. [10, 12, 13, 14, 15]

🌌 Navigate Art’s “Latent Space”

For creators and developers, the Gen Studio Project—built alongside Microsoft and MIT—uses Generative Adversarial Networks (GANs) to alter your viewing experience: [16, 17]

Visual Blending: A custom dashboard lets you manually adjust a multi-dimensional slider to morph, interpolate, and blend entirely different artistic styles or periods into completely new AI-generated visuals. [16, 17]
My Life, My Met: This tool parses personal Instagram uploads and matches your daily photo compositions with structurally identical open-access art pieces. [18]