Analogue generator - Personal discriminator
Pint of Science festival 2018 - Creative Reactions, created in response to the talk by Petar Veličković - AI: Generative Adversarial Networks Generative adversarial networks (GANs) are two computer neural networks trained together to first create data (the generator) and then attempt to distinguish whether any given data is real or generated (the discriminator). Generator/ If we can create data we can understand it better; so I generated my own fake synthetic data from analogue sources. Using monoprint noise and medium format film portraits I have created a lenticular visualisation of ‘latent space’, or the possibilities in between images. Discriminator/ Real and fake don’t necessarily have a binary relationship. What we believe to be real or fake, or somewhere between the two, is a reflection of who we are at that time - our beliefs, culture and training.