Storyboarding a Serious Game Environment for Evaluating the Impact of Empathetic AI for Sonar Operators: Empathetic AI for Sonar Operators.

Sonar operators work under high pressure and stressful conditions, handling large volumes of data that they must quickly perceive and interpret in order to make high-stakes decisions. Artificial Intelligence (AI) has the potential to aid them by processing large amounts of information and highlighting the critical data amongst the "noise" for the sonar operators to focus on. In order for such an AI to be effective, we need to build a cooperative human-AI team. Many studies show that such cooperation is built through trust [1] and that such trust can be built through a mutual, empathetic understanding of one another [2]. We argue, then, that for a human-AI team to be most effective, we need an empathetic AI to work alongside our sonar operators as this will foster a trusting team dynamic. Our aim is to build this trust and in turn, reduce the cognitive demand placed on sonar operators to a level that facilitates optimal performance in such high-stakes scenarios. In the context of this research, it is critical to acknowledge the principles of Responsible Research and Innovation (RRI). Factors that demand particular attention under this framework encompass potential unintended consequences of our empathetic AI, the phenomenon of over-trust in artificial intelligence (AI) systems, the possibility of AI making erroneous recommendations, the potential for misuse, and the development of strategies to mitigate these concerns. We propose to address these potential risks through a multifaceted approach. This includes the introduction of confidence reporting mechanisms, emphasizing the operator's role as the final arbiter of decision-making, referencing empirical findings on over-trust, and fostering transparency regarding the inherent limitations of AI.