30. Data Curation + AI: Principled Approaches for Ethical AI Design
Sarah Elaine Bratt, Syracuse University School of Information Studies
Posted: February 28, 2022 Accepted Languages: English/Inglés/Inglês
Data curation comes from the Latin root “to care”; yet, the data collections and systems have not reflected historically marginalized voices, people systematically made invisible across physical and digital platforms. We ask: How are AI systems, thinking, literacies, and individual and collective practices of data curation intersecting to amplify or suppress voices in the archive? How AI systems, computational thinking, and policies of access manifesting in archival work, broadly conceived? In this panel, we open new directions in questions about the ethics of AI in the context of data curation. We invite submissions revealing bellwethers for principled approaches to automated systems for data curation, archives, and data management. We will think together, and with, scholars and practitioners like Catherine D’Ignazio and Laura Kline (Data Feminism) about the ways to reconfigure data – via crowdsourcing, advocacy, and systems-building – to reunite, heal, and recuperate representations of ourselves and societies.
While this area has received a great deal of attention in recent years, it has been in mainstream spaces (Safiya Nobel’s Algorithms of Oppression, Future of Work @MSR, Sabina Leonelli’s critiques of “Big Data Collections”). We still lack robust discussion and principles approaches to AI in data curation. We welcome studies across industry, employee records, cultural heritage data, biometrics databases, and scientific data, especially the liminal spaces and emergent areas where policy is still “squishy” or absent. Our contribution to the field is a principled approach to ethics of AI in data curation, and methods for doing caring curation, e.g., participatory and speculative design.
