Frontiers in Data Privacy and Tech Ethics

A research talk given by the Society, Ethics & Technology team at the Orthogonal Research and Education Laboratory, presented at the New York Celebration of Women in Computing 2022. The session ran as a set of short segments, each team member taking a current question in data privacy or technology ethics:

• Bradly Alicea — A Collective Cognition Model for Ethical (and Sustainable) Open Source
• A. R. Ordis — Data Trusts for Data Privacy?
• Brian McCorkle — Explainability & Interpretability in Machine Learning
• N.T. Love — Log4j & Open Source
• Amaris Arias & Jesse Parent — Timnit Gebru’s DAIR + Big Tech News

The interpretability segment works through image synthesis as a case where a model can feel interpretable for anthropomorphic reasons, then sets Gary Marcus’s call for more generalizable cognitive models against Cynthia Rudin’s argument that no black box should be deployed for high-stakes decisions where an equally accurate interpretable model exists.

Recording: NYCWiC 2022 talk playlist