SODAS Data Discussion 4 (Spring 2026)
Copenhagen Center for Social Data Science (SODAS) aspirers to be a resource for all students and researchers at the Faculty of Social Sciences. We therefore invite researchers across the faculty to present ongoing research projects, project applications or just a loose idea that relates to the subject of social data science.
Two researchers will present their work. The rules are simple: Short research presentations of ten minutes are followed by twenty minutes of debate. No papers will be circulated beforehand, and the presentations cannot be longer than five slides.
Discussion 1
Presenter: Agnete Meldgaard Hansen
Title: AI, Ethics, and Care for Older Persons
Abstract:
This talk focuses on the ethical implications of the growing use of AI-based technologies in eldercare services (e.g. for needs-prediction, care-documentation and -planning, surveillance, and social interaction). Hopes are high that AI-technologies will enhance efficiency and save labor and thus help to counter the demographic challenges facing the welfare state. However, the use of these technologies may transform everyday care practices fundamentally, necessitating careful reflection on the values and ‘everyday ethics’ of eldercare with AI. Based on ethnographic fieldwork in the eldercare field and an ongoing literature review, the talk will outline key challenges and discussions related to ethical use of AI in care for older persons.
Discussion 2
Presenter: Jun Liu
Title: AI Benchmarks Gaze: A Reflexive Diagnostic Matrix for Critical Evaluation
Abstract:
AI benchmarks are conventionally treated as neutral measurement instruments—standardized tests that produce comparable, cumulative evidence of machine capability. This forthcoming paper argues otherwise. I reframe AI benchmarks as simultaneously technical and social artefacts: systems that do not simply record or measure a pre-existing world but negotiate what intelligence is allowed to mean in each of the social, institutional, and technical worlds they pass through. To make this negotiation visible and analytically tractable, I develop a Reflexive Diagnostic Matrix organized around various analytical dimensions. The matrix functions as an analytical probe: a structured set of questions designed to surface the assumptions, exclusions, and power relations that benchmark scores routinely render invisible. I demonstrate the matrix's application through case studies of selected dominant AI benchmarks. The Reflexive Diagnostic Matrix offers one of the first methodological foundations for that initiative.