AI for Preserving Cultural Heritage

I love the craft of art.

The materials, the smell, the feeling of creating with my hands, thinking through my hands. The challenge of learning to draw, learning to see so I can draw, to build mental models that translate three-dimensional reality into two dimensions. The contrast between what I envision and what the material allows. The learning curve, the countless techniques.

I want AI to amplify the value of physical craft, not obscure it.

When I train AI models on my own sculptures, the more digital iterations I create, the more precious the originals become. Photography had a similar effect on painting. The fact that so many people all over the world have seen endless reproductions of the Mona Lisa, movies featuring the Mona Lisa, artworks inspired by the Mona Lisa, all of this just makes the Mona Lisa the most known, and possibly the most beloved, painting for humans of our time.

‘Art in the age of mechanical reproduction’ is a cliché in digital art discussions. But the important point is: the original has an aura. I believe AI can help us highlight this aura. I've been developing a methodology that transforms my physical craft, writing, and images into foundations for AI-generated sci-fi worlds. This process could revolutionize how we preserve, share and continue traditional art forms.

Learning art history in the age of AI is a fun interactive practice.

In northeast Brazil, extraordinary sculptural traditions have evolved over centuries. Artists work with local materials, embedding stories, techniques, and rituals into every piece. Yet these traditions are increasingly marginalized. Television arrived, but producing TV content requires capital concentrated in Rio and São Paulo. International media, mostly American, also fills the gap. Local crafts rarely appear on screens.

But these sculptures are perfect source material for digital storytelling. They have so much narrative potential, rich aesthetics waiting to be animated!

Recently I visited the Pontal Museum in Rio de Janeiro. Every sculpture tells multiple stories, dense with details. Each could anchor its own film, or several. We could build entire worlds where these characters live and interact. Using just the museum's website photographs and AI video models (I recommend Eden.art , it's cloud based, no expensive equipment needed), I've created test animations, some of them with AI generated sound. The results are already compelling.

Until recently, film and animation production was prohibitively expensive and specialized. AI changes this. Animation is suddenly accessible to anyone with internet access (and there are some ways of getting free credits). As we all adapt to AI's presence, there’s also a push to amplify diverse aesthetics: Holly Herndon and Mat Dryhurst’s spawning, indigenous AI researchers like Jason Edward Lewis have been working on culturally-specific AI for years, the "Algorithmic Justice" movement addresses similar issues about whose aesthetics get encoded in AI, and Brazilian researchers like Giselle Beiguelman have written about decolonizing digital art and have offered important educational resources.

AI doesn't have to make everything look the same everywhere.

History doesn't move in straight lines. We've watched cultures converge, but the technologies enabling that convergence can now power divergence. Local crafts and traditions are more than heritage to preserve.

They're aesthetic languages ready to speak to global audiences in new ways.

When artisans see their work transformed into living animations, when their stories reach audiences they never imagined, when their craft is economically viable: that's a great use of technology. Humans like to create with their hands. We could aim towards economic sustainability through cultural authenticity. I want to see the craft artists being valued for the beauty they create.

Considering everyone is now forced to adapt to the world of AI, why not push some other aesthetics than Pixar and Disney?

Since communities worldwide must now navigate AI's presence, they should shape it with their own aesthetics and values. Right now, AI training datasets predominantly feature Western imagery. When local communities train AI on their own cultural productions, they're programming the future with their worldview.

What if every cultural museum had an AI animation station? What if schools taught AI through local craft traditions? What if we created open-source models trained specifically on regional art styles with licenses that pay back to the craft artists? The tools exist. There are so many great things to do!