Scholarlink

<p>As AI technology advances, human-AI interactions have become increasingly complex, exhibiting intelligent interactions. However, existing research primarily considers these interactions from a utilitarian perspective, seeing AI as a set of functional tools. In contrast to prior research, this study draws upon the human interactivity perspective and proposes various quality dimensions for the interactions between AI Voice Assistants (AIVAs) and users. The study also investigates which AI affordances enhance intelligent interaction qualities and how users’ characteristics affect the interactivity. We develop and test a model using survey data from 515 AIVA users with Partial Least Squares analysis. Our results confirm the salience of this new perspective and reveal distinct roles of interaction qualities, AI affordances, and user sociability in shaping intelligent interactions in determining the actual use of AIVAs. Our identification of the similarities and differences between human-AI and human-human interactions, based on different interaction qualities, provides insights for future studies and AIVAs design.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Why People Use AI Voice Assistants (AIVAs): Application of the Human Interactivity Perspective to Human-AI Interactions

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年2月28日

查看全文
  • Youngho Yoon,
  • One-Ki Daniel Lee

摘要

As AI technology advances, human-AI interactions have become increasingly complex, exhibiting intelligent interactions. However, existing research primarily considers these interactions from a utilitarian perspective, seeing AI as a set of functional tools. In contrast to prior research, this study draws upon the human interactivity perspective and proposes various quality dimensions for the interactions between AI Voice Assistants (AIVAs) and users. The study also investigates which AI affordances enhance intelligent interaction qualities and how users’ characteristics affect the interactivity. We develop and test a model using survey data from 515 AIVA users with Partial Least Squares analysis. Our results confirm the salience of this new perspective and reveal distinct roles of interaction qualities, AI affordances, and user sociability in shaping intelligent interactions in determining the actual use of AIVAs. Our identification of the similarities and differences between human-AI and human-human interactions, based on different interaction qualities, provides insights for future studies and AIVAs design.