Machine, Unlearning

Machine, Unlearning was created by a computer program that constructs “litanies” of questions by searching an automatically learned space for word clusters and contexts that are “near” a provided seed concept according to the machine-learned model. Ranging from obvious to awkward, natural to uncanny, these generated streams of inquiry work to frame the assumptions embedded within the model in a (somewhat) human-interpretable way. In the end, machine learning relies on data analysis and pattern recognition to achieve its goals. Machine, Unlearning, however, asks for a different type of analysis: examining, questioning, unlearning.