Adversarial Interactions and Poetic Disclosures: Artists as Agents in Explainable AI | AHRC CDA
Subject: Philosophy
HEI:Â University of Edinburgh
School:Â Edinburgh College of Art
Supervisors: Professor Ewa Luger, University of Edinburgh; Dr Siân Lindley, Microsoft Research; Dr Morgan Currie, University of Edinburgh; Beverley Hood, University of Edinburgh
Keywords:Â Artificial Intelligence, Research through Design, Human-Computer Interaction
Discipline+Catalyst: Creative Arts & Design
Knowledge Exchange Hub:Â Citizenship, Culture & Ethics, Creative Economies
Strategic Themes & Priority Areas: Creative Industries & Economies
About Martin’s Research:
This collaborative doctoral award project, in partnership with Microsoft Research, explores how adversarial interactions with ML and AI systems can contribute to both their development and the public’s understanding of them. By framing these interactions through the lens of agonistic democracy, this study emphasises the civic role of adversarial actions in revealing vulnerabilities and their logical limitations. With a focus on practice-based methods, such as Research through Design, this project proposes a new role for the artist and designer in interrogating these systems, contributing to AI explainability beyond notions of transparency.
CONNECT WITH MARTIN
Martin Disley
E: m.disley@ed.ac.uk
Â