World’s Fairs have always been about this idea of better living through technology. I’ve been fascinated by them for decades—visiting decommissioned sites in Shanghai, Seville, Montreal, Brussels, Osaka. Walking through these spaces after the crowds have gone, after the pavilions have emptied, you encounter something strange: the architecture of optimism in a state of suspension. These were time capsules, preserving the spirit of their era.
After Utopia began with a question: What would a World’s Fair look like if it existed only in virtual space? If you were to imagine a utopia today, it would likely exist virtually rather than physically. The metaverse, those “infinite, reconfigurable spaces” we heard so much about—this is where our collective dreaming has migrated. The nineteenth and twentieth centuries promised better living through technology. The twenty-first forces us to ask whether humans will retain control over what our life will look like.
I selected eighteen expositions spanning 135 years, from Paris 1889 to Osaka 2025, and began collaborating with The Wolfsonian-FIU to access their archives—the largest collection of physical blueprints, brochures, floor plans, and sculptures from pre-internet expositions. My team and I compiled imagery and artifacts, building a library that combined this historical material with digital archives, attendance statistics, and national anthems. We trained AI models on this comprehensive dataset, allowing the technology to participate in determining which pavilions to feature and how the architectural elements might behave.
The lines of people moving through the final work are algorithmically generated: they correspond to the actual number of visitors to each pavilion in its year of operation. This migration—people moving from one future to another—interested me because each Expo occupied a specific moment in time, a specific place on earth, yet promised something universal. In the installation, all of them exist simultaneously. Geography and chronology collapse into a single present.
AI became essential to this project, though not in the way people might expect. I used it extensively for research, for sketching, for processing vast amounts of historical information. But the dialogue between myself and the AI was a two-year process. It added a year to the schedule; I could have finished it a year earlier without this collaboration. That additional time was the cost of negotiating with a technology that can enhance our ability to dream—enabling us to conceive ideas larger and more fantastical than what an individual could imagine alone—while introducing an element of unpredictability, a shift in directions we may not fully control.
Growing up, I was inspired by the utopian architects of the 1960s and 1970s—Archigram, Superstudio. Their designs were driven by a concern for better lives, always conceived for realization in the physical world. The Plug-In City imagined mobile cultural hubs that could bring theaters and libraries to the countryside. AI allowed us to animate the buildings in After Utopia according to the movement of people, generating unexpected simulations that made each Expo come alive in ways those architects might have recognized.
The piece doesn’t editorialize. It doesn’t ask whether this is dystopia or utopia. I prefer to present the material without judgment. What I will say is this: as AI becomes an architect of experience itself, we risk losing not only control but imagination—the ability to define what kind of world we want to inhabit. The work distills all these time capsules into one hopeful future, just beyond reach, just over the time horizon. I think utopia is always out of reach. That’s what defines it.
After Utopia represents my attempt to use a technology that has the potential to both inspire and redirect our aspirations—to make a work that addresses how AI is becoming part of our world while asking whether we are still the authors of our collective dreaming or merely passengers in a cycle of autonomous imagination.
Marco Brambilla