From Explainability to Ineffability? ML Tarot and the possibility of inspiriting design
Caitlin Lustig, Daniela Rosner
DIS'22: ACM SIGCHI Conference on Designing Interactive Systems (DIS)
Session: Video Previews
Abstract
Explainability has become a dominant aspect of developing more accountable AI systems. But for AI to be more accountable, we as designers must also reflect on our own positioning—including aspects of ourselves that are difficult or impossible to fully explain, yet still influence our design processes. Just as designers develop explainable AI, we explore how AI can be used to develop the unexplainable designer through documenting our creation and use of a machine learning-generated tarot deck. Alongside design researchers, we consider the promise of such machine learning-generated artifacts for self-reflection on our creative and collaborative roles—and our responsibilities—when designing AI systems, and we discuss how the artifact sparked acts of inspiriting, a process of bringing the ineffable (back) into our engagements with machine learning systems.
DOI:: https://doi.org/10.1145/3532106.3533543
WEB:: https://dis.acm.org/2022/
30-second video previews of DIS 2022