Towards Contextualizing Embodiment of Intelligent Systems

Within human-computer interaction, a conversation has emerged regarding bias and empathy in intelligent systems. While existing research explores the potential for intelligent systems to exhibit these concepts, there has been less focus of operationalizing them. As the application of intelligent systems emerge in the health domain, this consideration of understanding the embodiment of these concepts becomes increasingly important. Through exploratory qualitative interviews, we examined experiences of bias and microaggressions within healthcare among 36 Black Americans and examined their perceptions of more ideal empathetic experiences. Our results reveal the constructs that impact Black Americans’ feelings of bias, including several prosadic factors of interaction with healthcare professionals that influence desired interactions with intelligent systems. Additionally, our findings identify the considerations the impact Black Americans’ perceptions of embodiment in intelligent systems. Our preliminary analysis provides insights for designers and engineers when developing dialogue and output of intelligent conversational systems.