Year: 2024
Type: article
Abstract: Situated at the intersection of language, discourse, and communication studies, the present study delves into the dynamics of human-artificial intelligence (AI) interactions. Our study centers on AI-based voice assistants which employ natural language processing to communicate with human users. With a dataset derived from 200 recorded interactions between human users and AI-based voice assistants of a leading Korean telecommunications provider, we investigate the intricate dialogue patterns that emerge within these exchanges. Employing the lens of conversation analysis, especially focusing on adjacency pairs, first pair-part (FPP) and second pair-part (SPP), our analysis elucidates how AI agents and human users negotiate meaning and interactional roles. We identify four distinct response types from the users’ SPP, revealing a variety of interactional patterns. The findings reveal that the users frequently respond to AI-initiated prompts with keywords, reflecting a strategy to efficiently retrieve information, and highlight instances of no verbal response. Additionally, the use of honorifics in Korean AI voice assistants underlines the influence of linguistic and cultural norms on the dynamics of human-AI interaction, emphasizing the need for AI systems to navigate social hierarchies effectively. Our study underscores the importance of enhancing human-AI dialogue and provides valuable implications for interdisciplinary research and practice in the rapidly evolving field of AI-based communication.
Source: Text and Talk
Language: English
Subfield: Language and Linguistics
Field: Arts and Humanities
Domain: Social Sciences
Open Access status: hybrid
Estimated APC paid: $2,489
OpenAlex ID: W4402877267