Machine Learning • Type • Web
‘Am I a Feminist Bot’? was a series of experiments that assessed the potential and challenges of machine learning tools to embody feminist practice through text and typography. It addressed the nature of bias, AI characters and personalities, and ethics regarding datasets, while also questioning the complexion of feminism and measuring its authenticity in automated contexts.
Embedded in the format of title, this project was led as a means of discovering and asking questions, answering some of them, and reserving the rest for rhetoric.
How can underrepresented genders, ethnicities, abilities relate to the condition of the machine in relation to society? Does a dataset based on fonts designed by women and underrepresented genders constitute a feminist dataset? How can I guarantee that the designers represented in the dataset hold feminist values? Is gender expression present in one's typographic practice? How do ‘girly’ and ‘feminine’ fonts differ from fonts designed by women? How can a machine learning model be tuned towards feminist creative practice? Are perfect and clean visual GAN outcomes anti-feminist? Is attempting to match machine ability to human ability anti-technofeminist? Is it actively feminist for a machine learning model to regurgitate feminist content that it was fed?