Pursuit of Coppélia

COPPÉLIA (1870) is a comic ballet based on short story DER SANDMANN by E.T.A Hoffman. The story follows a scientist who creates a beautiful mechanical doll, Coppélia, so lifelike that a whole village believes her to be real. Still popular today, the story plays upon our fear of the uncanny, that which appears human but is not.

Never has this discomfort been more than today. Cloning and artificial intelligence are accelerating at a rate so rapid that it is difficult to grasp the implications or consequences of what technology is capable of.

For many of us, sharing information and images of and about ourselves online is second nature. But who does this data belong to, and what can it be used for? The images shown here do not exist. With Artificial Intelligence, our online selves can live on far past our involvement with them. They can be pulled apart, rebuilt and made to do almost anything. For better, or for worse.

Using Machine Learning and Generative Adversarial Networks, this project in three parts explores the possibilities of creating new visuals and audio based on a person’s online footprint.

‘Pursuit of Coppélia’ is a 600 page hard back publication that follows the training process of a GAN learning to replicate my likeness.

‘Photographs I Didn’t Take’, taking photography work that I have posted across social media, I coded and taught a GAN to create images with similar characteristics, such as colour, composition and lighting.

‘Coppélia for the Modern Age’ is an audio visual piece exploring the latent space of AI images trained from clips of the ballet Coppélia, narrated by a clone of my voice.