File JMG-012 · Performance · AI · 2026
MODO NO SUPERVISADO
// Description
MODO NO SUPERVISADO is a real-time performance in which the audience sends a movement instruction and AI translates it through five languages back to Spanish, further each time from what was written.
Three people dance and interact with a device. The audience scans a QR code and writes, from their phone, a sentence that includes body movements with as many details as they like.
Each sentence goes through five randomly chosen languages and returns to Spanish: six versions in all. Every step is shown on the projection and heard in a voice of its own language, one after another; when the chain ends, a beep sounds.
The dancers move only to the audio. The piece shows how language and AI distort when translated in a chain.
- Year
- 2026
- Type
- Performance · real-time web · AI
- Duration
- 10 minutes
- Premiere
- 13 November 2026 · art gallery
- Languages
- Spanish → 5 random languages → Spanish
«Write a sentence that includes body movements, adding all the details you want.»
// Chain of translations
An instruction goes down six versions and distorts along the way. It illustrates the effect; it is not a real translation.
// Credits
- Concept, direction and development
- Juan Martinez Garcia