EID: Facilitating Explainable AI Design Discussions in Team-Based Settings

Artificial intelligence (AI) systems have many applications with tremendous current and future value to human society. As AI penetrate the aspects of everyday life, a pressing need arises explain their decision-making processes build trust familiarity among end users. In high-stakes fields such as healthcare self-driving cars, are required minimum standard for accuracy provide well-designed explanations output, especially when they impact life. Although techniques been developed make algorithms explainable in terms, no design methodologies that will allow software teams systematically draw out address explainability-related issues during conception established. response this gap, we proposed explainability (EID) methodological framework addressing problems systems. We explored literature on narrow down field into six major principles aid designers brainstorming around metrics guide critical thinking process. EID is step-by-step has refined over series user studies interviews experts explainability. It devised uncover resolve potential products simply refine explore The methodology novel aids stages pipeline can be integrated form card game. Empirical involving system shown decrease barrier entry time experience effectively well-informed decisions integrating solutions.