São Paulo 3024
Immersive 16-minute journey through São Paulo a thousand years ahead, using LLMs and LoRa as philosophical ground and bodily movement as live urban matter.
- Number
- 011
- Editorial status
- Felipe-authored work
- Reading regime
- system / body / mechanism
- Status
- REALIZED
- Year
- 2024
- Location
- Farol Santander São Paulo
- Context
- série dos Elementos / humano
- Type
- EXPOSICAO IMERSIVA
- Series
- Os Elementos
- Technique / medium
- LLMs · LoRa · OpenCV · projeção · instalação interativa
- Immersive room
- Seven-act, 16-minute journey through coexisting future worlds on the 23rd floor of Farol Santander
- AI
- LLMs and LoRa used as philosophical foundation rather than only image-generation tools
- Territory
- Interactive installation where body movements become architectural elements projected in real time
- Tracking
- OpenCV, five mapped projectors, and an RTX 4090 process bodily gesture as raw material for the urban landscape
- Infrastructure
- Eleven synchronized projectors, 5.1 sound, and a distributed three-zone network with redundancy
- Question
- Is technology illustrating the future, or asking whether the future is already hidden inside today's machines?
Public connections that situate this work inside the archive before the narrative text.
In 2024, I designed the entire 23rd floor of Farol Santander for a 16-minute journey to São Paulo’s future. Antonio Curti’s curation questioned what the city would be like a thousand years ahead — I wanted to answer with a technical question: if we used language models to imagine this future, would we be simulating realities or influencing the code of what’s yet to exist? I conceived the experience artistically and technically, working with LLMs and LoRa not as image generation tools, but as the philosophical foundation of the work. The journey passes through seven acts — floating islands in the sky, underground civilizations, fragmented remnants of the present — and the audience crosses luminous portals between coexisting worlds, not knowing if they’re seeing possible futures or multiple simulations of the same present.
Within the immersive experience, I created Território, an interactive installation where the audience’s body movements transform into architectural elements — buildings, cities — projected in real time by a system I developed with the AYA team. It’s not gamification: it’s the idea that inhabiting a space is building it. Each person who enters the room becomes an architect of a territory that never existed before and ceases to exist when they leave. I used OpenCV, five mapped projectors, and an RTX 4090 to process bodily gesture as raw material for the urban landscape. It’s an investigation of presence — not representation.
The entire project demanded robust infrastructure: eleven synchronized projectors for the immersive room, 5.1 sound system, distributed network in three zones with complete redundancy, and a team of 53 people in direct work, totaling 142 positions overall. I technically coordinated the assembly and executive direction, but what interests me to document here is the gesture of conception: designing a space where technology doesn’t illustrate the future — it asks if the future is already happening, hidden inside the machines we use today.