Oliver Haimson

@haimson.bsky.social

1813 Followers 1219 Following

Associate Professor at UM School of Information researching social computing, HCI, designing trans tech, social media, identity transitions, content moderation. 🏳️‍⚧️ he/him

Trans Technologies book cover
“AI may be everywhere, but it’s not for everyone - at least not yet. If we continue to ignore the perspectives of marginalized people, we risk building an AI-powered future that deepens inequities rather than reducing them.”
Cara and Mark singing
TIDEs: A Transgender and Nonbinary Community-Labeled Dataset and Model for Transphobia Identification in Digital Environments Authors: Francesca Lameiro, Lavinia Dunagan, Dallas Card, Eric Gilbert, Oliver Haimson
Alexis Shore Ingber presenting at the FAccT 2025 conference. Slide lists key takeaways: 1. People have significantly more negative attitudes toward emotion AI than AI. 2. People view all input data to emotion AI as sensitive. 3. Many fear negative consequences if their emotion data is shared, especially with powerful actors.
Bar graph showing average AI attitudes for participants in different identity categories. The plot shows that women (including both cisgender and trans women) have lower AI attitudes than men (including both cisgender and trans men); trans and nonbinary participants have significantly lower AI attitudes than cisgender participants; people of color, and Black participants in particular, have significantly higher AI attitudes than white participants; and disabled participants, specifically those who are neurodivergent or have mental health conditions, have significantly lower AI attitudes than participants without those disabilities. We used Bonferroni correction because we conducted multiple ttests, and we indicate statistical significance based on the following p-value thresholds: * p<0.005; ** p<0.001; *** p<0.0001