We made 25 years of a NYC gallery's exhibitions conversational with AI

Name: We made 25 years of a NYC gallery's exhibitions conversational with AI
Availability: InStock
Author: scriptedfantasy

by scriptedfantasy·Apr 9, 2026·1 point·0 comments

Visit Project View on HN

AI Analysis

●●SolidNiche GemSlick

In-gallery photo recognition makes this smarter than standard audio guide apps.

Strengths

•Bridges physical gallery visits with deep digital archive access seamlessly.
•Twenty-five years of unstructured exhibition data made queryable via voice.
•Clean interface that doesn't distract from the actual artwork experience.

Weaknesses

•Tightly coupled to one gallery's specific archive and physical space.
•Unclear if this scales to smaller venues without similar data infrastructure.

Post Description

We built Contxt, a conversational AI layer for art exhibitions. Today we're launching a public research preview with bitforms gallery in New York — 150+ exhibitions spanning 25 years of digital art, all accessible through chat and voice.

If you're in NYC, you can walk into bitforms, take a photo of an artwork from Maya Man's current show StarPower, and

Contxt will recognize it and let you ask questions about the piece, the artist's practice, and the exhibition.

If you're not in NYC, the full archive is available online. You can start with the current show and end up in a conversation about a 2003 exhibition. Everything you explore gets saved to a personal profile.

This is an early research preview — we'd love feedback on what works and what doesn't.

https://contxt.art/bitforms