Root Division, San Francisco, 2019

Engineering the Self is a video installation that presents the Self as a construction engineered through machine learning as mechanized by accelerated capitalism. A machine learning model called a Deep Convolutional Generative Adversarial Network (DCGAN) was used and trained on four separate datasets on images of human body parts. This model takes in large amounts of data in order to learn a hierarchy of representations. The videos depict what the model has learned, the interpolations being new and unique constructions generated by the model, both fascinating to watch and eerily real and unreal.

Using this real and unreal tension as an artistic metaphor, the companion essay looks at how machine learning models accelerate data collection as an ‘interactive mirror’, capturing and warping the individuals they process. The warped capture is multi-faceted, narrowing and filtering the many complexes of the self, and reflecting back certain, along with new, aspects of a person. The projection that exits leaves with traces of this journey, to mold the individual further in a cybernetic feedback loop.