“Hey Google,you’re stupid”: Framing conversational AI interactions in family discourse

Abstract

Given the growing popularity of AI tools in everyday life, recent research on conversational AI has highlighted the need to closely examine users’ interactions with AI in relation to the social settings in which they are embedded, so as to better understand what users do in and through such interactions. In this study, I analyze audio recordings of one family’s interactions with and around a voice-based Google device. Extending studies of the various ways in which frames are interrelated in talk, I demonstrate how family members reframe and blend frames to integrate their exchanges with Google Assistant into their interactions with each other, and show that evaluations of Google Assistant serve a key role in this management of frames. I also discuss how frequently attributions of intelligence and intent feature in such evaluations, which I suggest draw on and contribute to popular discourses of AI as “intelligent machines.”