Tangible Explainable AI - an Initial Conceptual Framework
Ashley Colley, Kaisa Väänänen, Jonna Häkkilä
Abstract
Artificial Intelligence (AI) solutions are becoming prevalent in almost all aspects of human life. However, their acceptance may be limited by a lack of transparency of how the AI works. Explainable AI (XAI) aims to provide the users of AI systems with an understanding of why decisions are made, increasing trust in the system. To date, research into XAI has focused on the use of graphical user interfaces, presenting numerical, textual or graphical explanations. However, AI is increasingly being used in systems that include physical devices, and hence the need for explainability in physical or tangible user interfaces (TUI) is also increasing. We present an initial conceptual framework for tangible explainable AI (TangXAI), which identifies the potential approaches of communicating XAI through physical artifacts, using the concepts of data physicalization and tangible interaction. The framework provides a basis into which ongoing research of tangible explainable AI can be mapped and related research gaps identified.