This project used machine learning algorithms to create a self-portrait that illustrates how we create our digital selves, how others are able to use our data to create a form of us, as well as the veracity and distortion of those creations. I primarily used data I created about myself and this data was fed into machine learning models whose outputs formed the self-portrait. Though self-portrait is the term used, my face per se is not necessary and the intention of the project was much more about the digital representation of human perception and experience. A self-portrait method is used to provide unrestricted, accessible, and emotionally vulnerable data based on a real person. Using myself as the primary subject reduces (though it does not wholly remove) problematic issues such as misrepresentation, external bias about the subject, and manipulation of the subject. The unique aspect of the project is in the use of multi-media, constructed out of definitive data, to portray a broad and equivocal interpretation of a person, which in turn calls into question how we create and are created by interpretations of data.

Untitled

A video summary of the work done for my MFA dissertation piece

Add

James Lawton