Algorithmic Stages: How AI Reinvents Live Performance

This book explores a quiet but profound shift in the performing arts: artificial intelligence is no longer content with generating images or texts offstage. It is stepping onto the stage. It performs. In seven chapters, the book first traces the history of performance art from Futurism and Dada to body art and digital environments in order to understand its essence: corporeal presence, ephemerality, risk, and the direct relationship with the audience. It then introduces the technical basics of AI that artists need (generative models, GANs, transformers, no-code tools) before analyzing three major functions of AI in performance: 1. Score generator : AI writes texts, movements, or scenarios in real time, making each performance unique. 2. Interpreter or co-performer : conversational agents, robots, or reactive systems dialogue and improvise with the human artist. 3. Participatory device : the audience influences the AI through their reactions, becoming training data themselves. An aesthetic and philosophical chapter deepens questions of authenticity, authorship, algorithmic error, and temporality. A final chapter offers practical guidelines for creating your own AI performance (tools, latency, documentation, copyright). The book does not take sides for or against AI. Instead, it shows that the tension between flesh and code, between human risk and algorithmic coldness, is precisely what makes these new art forms exciting. It invites artists and curious minds to experiment, with no programming skills required. AI is not a threat to live performance it is a strange partner, a non-human mirror, and perhaps the next great adventure of the stage.