Qingyi Ren, Basel Academy of Art and Design

Fall Semester 2021

Data is an essential part of machine learning, and the field of machine learning relies on the data to train, validate and test models. The detection and analysis of images is a related area where machines can 'see' a person's identity (Scheuerman et al., 2019). The study of gender classification through facial analysis techniques is at the heart of the discussion about the impact on AI on humans, with particular attention to race and gender bias.

This project will critique the contradictions between data classification methods and gender identity studies and develop a theoretical and practical framework for the implementation of 'critical digital gender narratives.' The political violence in artificial intelligence is transparent by discussing how data-led machine learning techniques suppress gender diversity and deepen gender bias. Rethinking how data is collected and categorized, combined with research on the intersectionality of gender perspectives and identity, suggests methodological possibilities for data processing to help build accountable and transparent AI.