Abstract
Artificial intelligence (AI)-related technologies available to consumers have been rapidly advancing but without parallel insight into the social process of their design. Building AI tools to converse with humans and support our lives remains a significant challenge for which consistency and codified standards are lacking. Following a tradition in science and technology studies of shining light on design and technology development as negotiated cultural processes, this article presents findings from ongoing interviews and fieldwork exploring the collaboration between engineers and user experience (UX) designers collaborating on conversational AI systems. To balance ethnographic data from a Tokyo-based conversational AI startup, interviews (n = 20) also include an international sample of people working for and with the world's largest technology companies. I describe how usability determinations unfold through daily inter-office negotiation and suggest that UX professionals in AI production primarily engage in two kinds of dialog: The first is human-centric and focused on the emotional needs and material conditions of human-computer interaction. The second is a technical dialog grounded in the parameters of what code “can do” (or what an individual engineer understands to be possible) and bridges the humanistic and technical perspectives necessary for AI system design.
Get full access to this article
View all access options for this article.
