How can we explain the broad and uneven spatial effects of Machine Learning (ML) algorithms that mediate the everyday lives of smart city residents? The discriminatory impacts of civic algorithms remain opaque to city inhabitants and experts alike. Current Explainable AI (XAI) approaches, while infuential, are limited in their ability to explain the inequitable algorithmic spatial efects in an accessible, critical, and grounded manner. My thesis explores the potential of participatory mapping as a critical and collaborative technique to address these limits. My work draws on (1) scholarship on critical data and algorithmic studies, (2) qualitative research with domain experts from history and criminology, and (3) participatory mapping sessions with city residents and ML practitioners. Ultimately, my research will inform the design of a toolkit to help people in classrooms and community centers collaboratively refect on how city residents may unevenly experience the impact of artifcially intelligent systems guiding contemporary urban life.
1. Making Smart Cities Explainable: What XAI Can Learn from the “Ghost Map”. Shubhangi Gupta, Yanni Loukissas. Late Breaking Works paper at CHI 2023.
2. Mapping the Smart City: Participatory approaches to XAI. Shubhangi Gupta. Doctoral Consortium at Designing Interactive Systems (DIS) 2023.