(2021) Single Channel HD video,
1 minute 50 seconds
‘Duel’ is conceived as a fight between images, a struggle to produce meaning within opaque, automated systems of interpretation, and an increasingly competitive cultural arena. The work is a short meditation on how subjectivity is classified in digital media environments wherein images and sounds may be produced and interpreted largely by non-human actors.
The work reflects upon the use of General Adversarial Networks (GANs). In the context of machine vision, GANs are a kind of neural network used to analyse and generate images. GANs are used both literally, in the production of the work, and metaphorically, to reflect upon the linkages between quantification, word and image, sign and referent. Two voices are produced with AI audio software that produces synthetic voices replete with the ability to emote: users can have the voices perform a script as if they are happy, sad, scared etcetera. In a scenario located somewhere between a performance review and therapy session, two characters discuss an elusive goal. With image as weaponry, word as shield, a semiotic contest is followed by a descent into gobbledygook.
This work was exhibited in the Group exhibition ‘Image Collective’
at Blindside in 2021