STROLLING CITIES

The project focuses on nine Italian cities—Milan, Como, Bergamo, Venice, Bologna, Genoa, Rome, Catania, and Palermo—and draws from millions of photographs taken by students of the Politecnico di Milano's AUIC School during the 2020–2021 pandemic lockdowns. These photographs deliberately avoid iconic landmarks, instead capturing the raw texture of daily life: walls, streets, windows, facades, and ordinary objects that make up the urban fabric. What emerges is an alternative portrait of the city—one stripped of its stereotypical imagery and returned to the presence of its material reality.

A custom generative AI model, created by Mauro Martino and Luca Stornaiuolo, was trained on this dataset. The AI transforms static photographs into continuous, flowing video paintings, generating ambiguous and mutable cityscapes. These moving images dissolve clear boundaries and suggest a city in constant flux, freed from predefined roles and open to infinite reinterpretations.

This perceptual shift is further amplified by the Voice-to-City system, an interactive interface that enables users to modify the AI's visual output using their voice. Spoken input—ranging from single words to entire poetic verses—shapes the way the urban landscape appears on screen. Poetry was chosen as the primary input form due to its condensed sensory power and its deep link to both orality and literacy. The use of poetic language produces unexpected visual outcomes, often surreal and estranging, that challenge the viewer's expectations and expand their mental image of the city.

Interestingly, because the AI is trained to identify patterns rather than unique elements, it tends to generate representations that emphasise repetition, such as ordinary streets, recurring architectural details, shop windows, and everyday city corners. While iconic monuments fade into the background, these more subtle features emerge with a newfound intensity and beauty. For those who know these cities intimately, the experience evokes a quiet recognition—an emotional, non-touristic sense of place.

To complement and enrich this exploration, excerpts from Invisible Cities by Italo Calvino are woven into the experience. Calvino's poetic vision of cities as metaphors and shifting mental constructs aligns perfectly with the project's intention: to depict urban space not as a fixed entity but as a living, reconfigurable field of imagination and memory.

In this way, Strolling Cities becomes a poetic, technological stroll through space and language—an invitation to reflect on the layered complexity of our urban environments and to rediscover beauty and meaning in their most overlooked elements.

Video: https://vimeo.com/553601323 

Learn more and play with the interactive module at the strollingcities.com

 

 Credits
 Mauro Martino – MIT IBM Lab, AI Artist
 Ingrid Maria Paoletti – PoliMI, Dip. ABC, Material Balance Research Coordinator
 Maria Anishchenko – PoliMI, Dip. ABC, Material Balance Research
 Seyma Adali – PoliMI, Dip. ABC, Material Balance Research
 Luca Stornaiuolo – Generative Models
 Dalila Colucci – Selection of Poetry
 Emanuele Strano – MindEarth

Narrated by
Federica Fracassi
Michele Di Mauro

Journey Team
 Begum Sardan, Esra Kagitci, Giulia Annese, Giulia Grassi, Gokhan Dede, Guijia Zhao, Guitsa Herro, Huihong Zhai, Ilaria Donadel, Luca Spolaore, Manuel Zangirolami, Martina Ghidini, Michelle Rodriguez, Olga Beatrice Carcassi, Paoloandrea Paganotto, Sebastiano Nespoli, Shangyu Lou, Yinglei Li.
 

In collaboration with
 Politecnico di Milano – Material Balance Research
 MIT – IBM Watson AI Lab
 MindEarth