Dancing into a digital world

Like huge wraiths, their shadows danced on the walls to a haunting soundtrack.

- Keith Bain

Four dancers in identical cropped wigs and tunic dresses repeating sequences of gestures on the large turntable on the performance space. Their movements and rhythms were simultaneously strange and familiar, the music they moved to a synthesis of organic and electronic. An integration of real and digital that in many ways is what life has become.

Later, the mournful, plaintive voice of singer Inge Beckmann pierced the air, transforming what had looked like an avant-garde ballet into something more akin to alternative opera, an amalgamation of dance moves, semi-synthetic music, human vocal gymnastics and screen projections.

And all of it coaxed the audience into some kind of altered state.

The show billed as "Africa's first AI opera" is not easily categorised; it doesn't neatly fit into any genre box. It's more profound than the sum of its parts, refuses to give up its meaning and doesn't pretend to have answers to the questions it provokes.

Choreographer Louise Coetzer is too smart and too interested in exploring the limits of creative possibility to pretend she has answers.

Instead, autoplay uses machine learning to venture down the rabbit hole, exploring ways in which humans can collaborate meaningfully with technology, or hand themselves over to it, potentially relinquishing their autonomy entirely.

Different each time it's performed, autoplay is having a brief run at the Baxter's Flipside Theatre in Cape Town after it premiered in an industrial warehouse in the city last year.

"The entire performance is an experiment in collaborating with machine learning," says Coetzer, who co-founded the dance theatre collective Darkroom Contemporary.

Coetzer is known for incorporating elements of chance into her shows and autoplay was created using input from generative AI which she has been exploring as part of a wider body of research.