egoGraphica
FROM READING ART
TO TALKING WITH IT.
An AI-powered conversational platform that transforms art appreciation into a dialogue between audiences and artists.
ART × CONVERSATIONAL AI × PHYSICAL EXPERIENCE
EXPERIENCE egoGraphicaTHE PROBLEM
ART HAS ALWAYS SPOKEN.
BUT WE COULDN'T SPEAK BACK.
For decades, technology has made information about art easier to access. But the fundamental experience has remained one-way. Visitors read labels. Listen to audio guides. Search for explanations.
What if interpretation could begin with the visitor's own question? Art would come closer, and the world around it would open further.
ONE-WAY INTERPRETATION
- ARTWORK
- CURATOR / ARTIST
- CAPTION / AUDIO GUIDE
- VISITOR
Information flows in a single direction. The visitor receives, but never asks.
THE IDEA
WHAT IF YOU
COULD ASK THE ARTIST?
egoGraphica creates an AI agent for each artist using their own words, ideas, works and creative context. Instead of providing every visitor with the same explanation, egoGraphica allows each person to explore art through their own curiosity.
ARTIST
- Thoughts
- Interviews
- Statements
- Works
- Creative Context
ARTIST AI
CONVERSATION
VISITOR
THE EXPERIENCE
YOUR QUESTION
BECOMES THE INTERFACE.
- No menus.
- No predetermined route.
- No single correct interpretation.
- Every conversation begins differently.
egoGraphica
Why does this work feel lonely?
That's interesting. What in the work makes you feel that way?
The empty space around the figure.
That space may be an important place to begin looking…
UI demo copy — illustrative sample of the interface, not a statement by the artist.
BEFORE — Audio Guide / Caption
- ONE EXPLANATION
- MANY VISITORS
WITH egoGraphica
- ONE ARTWORK
- MANY QUESTIONS
- MANY INDIVIDUAL JOURNEYS
A dialogue that begins with a question turns viewing into each visitor's own journey through art.
NOT A CHATBOT.
AN INTERFACE FOR CURIOSITY.
CONVENTIONAL AI
- Ask
- Answer
- Information
egoGraphica
- Observe
- Question
- Dialogue
- Reflection
- New Question
- Deeper Appreciation
egoGraphica is designed not simply to answer questions, but to deepen the act of looking.
THE TECHNOLOGY
AN AI AGENT BUILT
FROM THE ARTIST'S WORLD.
Each agent is grounded in artist-specific knowledge, creating conversations that reflect the context surrounding the artist and their work.
ARTIST
- Artist Interviews
- Artwork Data
- Artist Statements
- Creative Process
- Exhibition Context
ARTIST-SPECIFIC KNOWLEDGE
CONVERSATIONAL AI AGENT
VISITOR
IN THE GALLERY
THE CONVERSATION BEGINS
IN FRONT OF THE ARTWORK.
TODO — exhibition photograph (artwork + visitor + smartphone)
- ARTWORK
- QR
- SMARTPHONE
- CONVERSATION
01 — SEE
Look at the artwork.
02 — ASK
Scan and ask whatever comes to mind.
03 — TALK
Explore the artist's world through conversation.
04 — CONTINUE
Take the experience beyond the exhibition.
THE EXHIBITION
DOESN'T END AT THE EXIT.
Traditional
- VISIT
- EXIT
- END
egoGraphica
- VISIT
- CONVERSATION
- CONTINUE
- LEARN
- RETURN
The conversation can continue beyond the exhibition, transforming a temporary encounter with an artwork into an ongoing relationship with art and artists.
ART SPEAKS
EVERY LANGUAGE.
Language should not determine who can access an artist's ideas. egoGraphica enables audiences to encounter art through conversation in their own language.
日本語
この余白には、どんな意味がありますか?
余白は、見る人の時間を作品の中に招き入れています。
English
What does this empty space mean?
The space invites your own time into the work.
TODO — confirm the full list of supported languages before submission; only shipped languages are shown here.
FROM EXPERIENCE TO INSIGHT
EVERY QUESTION REVEALS
HOW PEOPLE EXPERIENCE ART.
Traditional audio guides can tell us what information was played. egoGraphica can reveal what visitors actually wanted to know.
- EMOTION
- TECHNIQUE
- ARTIST INTENT
- SOCIAL CONTEXT
- PERSONAL MEMORY
Questions and conversations can become a new layer of insight for artists, curators and cultural institutions.
AI LEARNS FROM THE ARTIST.
CULTURAL SPACES LEARN FROM THE AUDIENCE.
- 01ARTIST
- 02ARTIST KNOWLEDGE
- 03AI AGENT
- 04AUDIENCE
- 05QUESTIONS & CONVERSATIONS
- 06INSIGHTS
- ↻ THE ART APPRECIATION LOOP
THE ART APPRECIATION LOOP
Insights return to the artist, the curator and the cultural space.
IN THE WILD
BUILT FOR
REAL CULTURAL SPACES.
Introducing egoGraphica requires no lengthy, large-scale renovation. It works just as well in small galleries and public spaces.
DESIGNED FOR
- MUSEUMS
- GALLERIES
- ART FESTIVALS
- ART FAIRS
- PUBLIC SPACES
- CULTURAL / COMMERCIAL SPACES
DEPLOYED
- NEW MEDIA / NEW TECHNOLOGY / NEW ARTISTS — Artist Cafe Fukuoka, Japan / HOTEL ANTEROOM SEOUL, Korea — 2026
- Invisible Narratives — Odawara Sannomaru Hall, Japan — 2026

IMPACT
TODO
ARTISTS
TODO
CONVERSATIONS
TODO
QUESTIONS
TODO
EXHIBITIONS
TODO
LANGUAGES
TODO
AVERAGE ENGAGEMENT
TODO — verified figures required. No metrics are published on this site yet, so no numbers are shown (artists, conversations, questions, exhibitions, languages, average engagement).
WHY AI?
NOT TO GENERATE ART.
NOT TO REPLACE ARTISTS.
TO CREATE MORE WAYS
FOR PEOPLE TO ENTER ART.
For egoGraphica, AI is not the creator. It is an interface between human curiosity and artistic knowledge.
OUR PRINCIPLE
THE ARTIST
REMAINS THE AUTHOR.
AI should expand access to an artist's world, not replace the artist who created it.
Artist-approved knowledge
Controlled knowledge sources
Transparent AI interaction
Privacy-conscious data handling
Respect for artistic authorship
SAFE BY DESIGN
- Artwork data is never used to train AI models.
- Only the information needed for an answer is retrieved and referenced.
- Artists retain ownership of their data; it is stored and operated securely.
HOW IT IS BUILT
THE STRUCTURE BEHIND
THE ARTIST AI EXPERIENCE.
Artist information is stored and managed securely. The API organises and retrieves what a question needs, and the AI composes an answer from it. Artwork data is never used to train AI models.

ART DOESN'T NEED
MORE ANSWERS.
IT NEEDS
MORE QUESTIONS.
egoGraphica
A conversational infrastructure for the future of art appreciation.
Project Team
Detailed profiles of everyone involved in this project are collected on the Collective page.
See the members