Human-AI Collaboration: From Explainable AI to Co-Creating Meaning

As AI is more commonly used in decision-making, it has become increasingly important to understand how humans and AI systems can collaborate effectively. One of the aspects of human-AI collaboration is explainable AI (XAI), which aims to make decisions made by AI systems comprehensible to humans. This paper introduces a framing perspective on XAI in human-AI collaboration. With this theory, we propose a shift from persuasion-based XAI, where humans passively receive algorithmic explanations, to co-creation in XAI, where humans and AI collaboratively create meaningful explanations. We motivate this shift through a frame analysis contrasting conventional XAI development with empirical data from views on XAI in sleep medicine. The paper contributes new insights for enhancing human-AI collaboration through a more interactive and contextually aware approach to XAI.