unit-code
UtopAI is a game that examines the relationship between historical narratives and contemporary Artificial Intelligence in the design of cities. The history of utopian design is hundreds of years old. By imagining alternative societies, these worlds held a mirror up to reality. Rather than using technologies like AI to make cities more ‘efficient’, UtopAI creates a world where players can travel to historic utopias and train AI in their social and cultural principles, providing learnings that can then be applied back into the real world. The player’s AI functions as both a sidekick and a history generator, learning and responding to the decisions made in the game. Through repeated learning, the AI grows from a ‘baby AI’ to an ‘adult’ state, enabling it to create new tools for the player to use in the design of their real-world city. By combining human-led decision making with AI-led analysis, the game allows players to create a complex picture of historical utopia, literally researching their culture and meaning together. In this way, UtopAI suggests that future cities, and the AI that will undoubtedly organise them, can learn from new types of information – the fictional worlds that have helped to define our collective cultural understanding of space.
By training their AI, players can unlock new design tools and principles to apply back in their real-world home city.
The game is built around seminal utopian projects from the past including Thomas More’s Utopia, Le Corbusier’s Ville Radieuse, Archigram’s Walking City and Tao Yuanming’s Peach Blossom Spring.
As players explore the various utopian worlds, their partner AI will record architecture and culture, from building types to flora and fauna.
The game offers a variety of ‘home cities’ for players to redesign including Brasilia, Osaka and Oslo. Each city’s form reacts to the utopian principles in different ways.
Here we can see the partner AI learning the architectural form of a house in Tao Yuanming’s Peach Blossom Spring.
As the AI assimilates more information based on the player’s actions, it grows from a ‘baby AI’ to an adult version, in turn offering the player more tools to use.
Key architectural forms from the utopian cities can be directly learned and reconstituted in the player’s home city.
Players explore and learn utopias in a free-roaming third-person interface before reverting to a ‘city-builder’ style game where they can grow and organise their new constructions.
As players change their real-world cities with utopian principles, they can obtain live feedback from citizens to the changes made.
Players will ultimately travel back and forth between building their cities and journeying to train their AI in different ways by visiting utopias. These visits can also cause the city to change as the AI learns habits.
Various new building components are unlocked by training the AI in a utopia. These then serve as building blocks to augment the existing city.
Here we can see various city forms produced by the combination of player and AI decision making.
The player’s AI partner will also read landscape forms and use these to automatically generate new contextual settings for the player’s home city based on their exploration of utopias.
At the end of the game, players will have generated a city through cooperation with an AI that embodies the architecture, culture and politics of utopian urbanism combined with advanced computation.