Three Stage Drawing Transfer
Symposium:
Title:
- Three Stage Drawing Transfer
Organiser/Presenter(s):
Statement:
This work creates a visual-mental-physical circuit between a generative adversarial network (GAN), a co-robotic arm, and a five-year-old child. From training images to the latent space of a GAN, to a live human interpreter, it establishes multiple translational stages between human and non-human actors playing out through the medium of drawing.
The title refers to Dennis Oppenheim’s 1971 performance ‘Two Stage Transfer Drawing’, a direct inspiration for this work. In that piece, Oppenheim staged a photographic performance with his son, reconceiving drawing as a form of intimate, embodied, inter-generational touch-based communication. I have added additional stages to the exchange: from the host of absent children contributing training images to the blackbox neural network (GAN); through the robotic arm drawing with the child; to my son’s eyes where they are seen, named, and rendered through his own hand and imagination.
The neural network at the center of this artwork is a StyleGAN2 architecture trained on a collection of over 7,000 children’s drawings gathered by Rhoda Kellogg. As visual subject matter, these “early graphic expressions” recall Claude Debuffet’s Art Brut (1948) or raw art. I am interested in children’s art not just for being outside of convention and learned expression, but for how it shows drawing and graphical expression in its moments of development. Perhaps a better reference for this endeavor are the Surrealists, who also sought to escape the personal bounds of language and subjectivity through sleep deprivation, mind-altering substances, and games of chance and collaboration.
These questions of where we seek the other; where we grant agency, autonomy and intelligence; and why we might wish to escape our own subjectivities speak to the design and use of emergent AI. Experimental interactions between humans and non-humans have the potential to produce mutual revelation, this project is one model for that.
