“See in CMYK” by Stefanie Posavec combines the 100-year-old printing press with AI, in collaboration with the Exploratorium and Google Arts & Culture.

Your browser does not support the audio element.

The Exploratorium in San Francisco is a place where art, science and human perception come together. Artists are invited to reimagine scientific phenomena in ways that spark curiosity and experimentation, creating experiences that help people see familiar things from new perspectives. Stefanie Posavec’s latest digital artwork, “See in CMYK,” does just that.

The Exploratorium and Google Arts & Culture invited Posavec to create “See in CMYK” as an opportunity to bridge artistic imagination, scientific curiosity and emerging technology. Collaborations like this create new ways for people to engage with ideas that sit at the intersection of art and science, moving them beyond observation and into active experimentation.

"See in CMYK" builds on Stefanie Posavec’s physical mural, "A Four-Color Field,” originally commissioned for the Exploratorium’s Surface Space mural series.

For over a century, the standard color printing process has used just four basic inks — cyan, magenta, yellow and key (black) — to trick our brains into seeing millions of vibrant colors. Up close, Stefanie's mural is an intricate collection of tiny, colorful symbols: cyan clouds, magenta lips, yellow suns, black cats. But from afar, the experience transforms. Viewed from a distance, these thousands of isolated icons melt together to create a singular large image of iconic California poppies.

Stefanie Posavec’s "A Four-Color Field" installation at the Exploratorium. Credit: Ida Tietgen Høyrup for the Exploratorium

How AI transforms color into play

"See in CMYK" brings the physics of traditional print processes that "A Four-Color Field” explores into the modern day with help from Google’s Gemini 3 Pro Image.

The digital interactive artwork invites users to take a photo, and the Gemini AI model transforms it into cyan, magenta, yellow and black icons that people can interact with and manipulate.

Traditional image processors can mechanically separate photos into raw color layers, but they cannot understand what they are looking at. By using Gemini, “See in CMYK” semantically analyzes your uploaded photo to curate a bespoke universe of icons mapped directly to the content of the images. It brings together the mathematics of color with a personalized artistic experience.

This camera experiment emulates the traditional printing process with Gemini.

How it works

This interactive tool relies on a simple recipe of four inks that work together to challenge human perception. Here is how you can create your own:

• Choose your canvas: Take, select or upload a photo from your device. Gemini identifies the subjects and light levels within your image.
• Transform with AI: Watch as traditional, uniform halftone dots are transformed using a library of 4,000 pre-generated AI-icons to match your photo’s content.
• Learn by play: The tool is configured to support learning through experimentation. Adjust the settings to see immediate changes, keep tweaking until you’re happy with your transformation — then generate your own unique four-color poster designed by Stefanie Posavec.

Customize, download and print your poster

Ready to see the world in four colors? Explore “See in CMYK” on Google Arts & Culture, experience it online with the Exploratorium or visit the Exploratorium in San Francisco this summer for a limited-time “See in CMYK” to play with a physical installation of the experiment in person.