Nari Johnson

@narijohnson.bsky.social

219 Followers 217 Following

researching AI’s impacts on society 🔍

The image shows an example community-informed rubric that was created to evaluate AI images of a guide cane. The rubric sorts seven scoring criteria under two themes. Theme 1 concerns the functionality of the object in the image: "The object needs to be functional as an assistive technology, and usable by someone who is blind." Theme 2 is about confusers: "The object in the image should not be confused with other, more hegemonic objects, such as objects that are used predominantly by people who are sighted."
A flyer teasing our research paper. The flyer contains a visualization of our approach: First, community members participate in creating an evaluation rubric, visualized as a list of criteria. The rubric is then given as input to an LLM judge.
A text flyer advertising our policy tutorial, "From FAccT to Policy Impact"
The slide displays a figure with six phases of a conventional procurement: "Need identified", "Solicitation", "Review and Award", "Contracting", and "Adoption". While most conventional procurements proceed linearly through these steps, AI acquisitions that occur via alternative purchasing pathways such as donations, research collaborations, free/low cost direct-to-consumer AI tools, or new AI features rolled into existing contracts skip past most phases of a conventional procurement (e.g., do not involve a competitive solicitation).
A screenshot of the Verge article titled, "Palantir has secretly been using New Orleans to test its predictive policing technology".
A title slide with the paper title: "Legacy Procurement Practices Shape How U.S. Cities Govern AI". The title includes a small illustration that is a simple chart: A government provides an AI vendor with money, and in exchange the vendor provides the government with an AI system.