By Cansu Peker

When I spoke with curator of the AI HOKUSAI project, art manager and producer, and CEO of TAtchers’ Art Management, Anna Shvets, she described a project that feels part laboratory, part homage, and part provocation. For more than a year and a half the initiative has been asking a deceptively simple question: what happens when you bring Hokusai’s way of seeing into a world shaped by code?

AI HOKUSAI is a research-and-residency project that pairs Hokusai’s radical, tutorial-style approach to making art with the latest generative tools. The goal is to use his spirit of curiosity and his appetite for experimentation, myth, and found objects as a test case for thinking through authorship, perception, and the role of technology in creative practice. It deliberately blurs the old/new binary: brushstrokes and code are treated as a hybrid language rather than as enemies.

How the project unfolded

The idea began in 2022 as a way of using AI to preserve and explore cultural legacy — to “go back in time” while asking what that backward glance means for the present. In 2024 the organizers ran an open call inviting artists to train neural networks on Hokusai’s work, and shortly after opened a residency/lab call. Early on they also launched an AI Hokusai Challenge: roughly 150 artists submitted experimental proposals, and three winners were selected and awarded prize money — an early demonstration of how AI processes can be used responsibly by crediting the historical source material.

From more than 400 applicants across 60+ countries, the program narrowed to a small international cohort of seven artists who participated in the residency phase. The residents who work across visual art, sound design, performance and traditional artisanship began collaborating with technologists and researchers across academia, business, and the social sciences. The residency kicked off in October 2024; by January 2025 the cohort had produced the artworks that would later be shown publicly in May 2025.

Anna made a point of telling me how practical the project was from day one: the organizers paid artists so they could focus on research and making, and they invested in the AI tools each participant wanted to use. For many artists, this was their first serious encounter with generative tools; for others it was a chance to test long-held questions about legacy, craft and new media. Across the board, artists helped each other: sharing tools, workflows and workarounds in a very collaborative remote environment.

Research, ethics and public programming

AI HOKUSAI positioned itself as research as much as exhibition. While the artists explored generative tools to make final works, the team simultaneously observed the process through cultural, technological, anthropological and philosophical lenses. The research will continue with an online survey of art professionals including artists, curators, managers, and critics about how they use and experience AI tools. The eventual research paper will compile those experiences to give a clearer picture of what AI means for art, culture and living heritage.

Public-facing programming ran alongside the residency: free webinars, discussions about ethics and aesthetics, sessions on AI for virtual spaces and even practical guidance on using AI in art business contexts. These were designed to provoke rather than settling debates and to bring wider publics into the conversation.

“Katsushika Hokusai, (1760–1849) was a worldrenowned Japanese artist who sought to bridge Eastern and Western traditions. He created instructional manuals called manga that have inspired contemporary art education through a systematic methodology and approach that algorithmically breaks down complex artistic processes into accessible steps and elements.

Hokusai’s influence extends beyond traditional Japanese art, permeating Western art and pop culture. His innovative use of perspective and composition influenced the Impressionists, with artists like Claude Monet and Vincent van Gogh studying his prints.

But his influence does not end there: Some claim that without Hokusai, various artistic disciplines such as modern woodcutting, graphic design, comics, manga, anime and even tattooing would not be the same.

The project organizers are confident that its methodology will make it possible to train neural networks to use the legacy of the great Japanese master as a tool for modern artists.”

Technical realities and unexpected discoveries

The residency also surfaced real technical limits. For Tomo Sone, a resident working between performance and digital tech, the producers took a 360° capture of the Japanese dancer and tried to turn her choreography into a 3D sculpture. Integrating that sculpture into the project’s virtual environment (built in Unreal Engine) proved painstaking.

The team didn’t want a game-like feel; they wanted a space that read as believable and gallery-like. To reach that aesthetic they partnered with the French architecture firm Bechu & Associes to finesse lighting, scale, and atmosphere. They also made a point of accessibility: the virtual space works on desktop, smartphones and VR headsets — a choice that complicated production but widened audiences.

Other creative problem-solving cropped up too. Sound designer Julio César Palacio used AI to imagine how instruments from Hokusai’s time might have sounded — some of those instruments no longer exist, and AI became a way of reconstructing sonic possibilities. Another artist, Alex May, asked the organizers to train the network on photos of everyday objects; the same sort of found items Hokusai collected, so he could translate that impulse into a 21st-century practice.

Because participants were scattered around the world, all collaboration happened over video calls. Anna was candid about the challenge, but also about the generosity that emerged: artists exchanging tools and tips, and a collective willingness to learn.

Where this goes next

By May 2025 the work from the residency had been presented to the public and met with a lot of positive feedback. The team plans to keep researching: interviewing more artists, talking with experts, publishing the survey findings in a report slated for October, and applying for nonprofit registration in the United States to sustain future cohorts and projects.

They’re also taking the project out of purely virtual spaces. Anna told me they’re lined up to present at REACT 2025 in Vienna just before the Culttech Summit, and they plan to show some of the works in their physical space as well.

Why this matters

“AI is here. It’s not going anywhere. Let’s figure it out,” Anna told me — blunt, practical, and optimistic. That, I think, captures the project’s tone. AI HOKUSAI refuses to treat today's tools as either miracle or menace. Instead, it opens a testing ground where heritage, pedagogy and machine learning can be examined together: to preserve, to teach, to create income opportunities for artists, and to spot where ethical, legal, and aesthetic questions must be asked.

“Art isn’t math; it doesn’t operate on formulas. It’s a space for trying things, new endeavors, and new technologies. Artists are crazy about exploring boundaries; they take what they’re given to test its limits and go beyond. So let’s give the toys to those who can expand them.” I love Anna’s playground approach to art!

AI HOKUSAI creates a place for artists and publics to explore, test, and argue — they connect people from different geographies and times. While connecting dimensions, they help shape how we steward cultural heritage into the coming decades.