.png){kind=icon}
.png){kind=icon}
Return
Someone Behind This Technology: Conversation Designers in the Artificial Intelligence Loop
.png){kind=icon}
PRE
AI
DOI：10.1080/19378629.2025.2575366.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Artificial intelligence (AI) tools are advancing rapidly, yet their development remains opaque, particularly in terms of the collaborative practices that shape their design. This article presents findings from ethnographic fieldwork conducted at a Tokyo-based conversational AI startup, supplemented by interviews with conversation designers and engineers (n = 20) working across major global tech companies. It examines how usability determinations emerge through daily negotiation between these actors and identifies two recurring modes of dialog navigated by conversation design professionals: one focused on the emotional and pragmatic needs of users, and another grounded in engineering constraints and assumptions about technical feasibility. Many conversation design and UX professionals hold academic training in the social sciences, and their work requires the ongoing translation of humanistic concepts into engineering language. While current conversation design remains largely human-authored, the balance between scripted dialog and generative AI continues to evolve. I argue that conversation designers' social science expertise is central to how AI systems come to feel 'human.' Bridging literatures in user experience (UX), human-computer interaction (HCI), and science and technology studies (STS), this article addresses gaps between humanities/social science and engineering research cultures, advocating for more integrative scholarship that reflects the collaborative realities of AI development.
Keywords:
LANGUAGE