My thesis explored ideas around non-linearity within the time and space of a narrative, discussing how the introduction of new concepts in Machine Learning might affect future outcomes. Download the text here.

The practical documentation covers the technical aspects of the project including my experimentation with Rebecca Fiebrink’s Wekinator. This involved creating a database of illustrated animation that the Machine Learning system then accessed. Using screen vision the camera detected the audience and their preferences in terms of narrative progression. See some images showing this interaction here.