Abstract

As AI is increasingly embedded in education, a pressing question emerges: Can AI instructors genuinely foster meaningful learning, or are we just mimicking knowledge? This study addresses this challenge by exploring how the acoustic design of AI voices influences human learning performance. Amid concerns that AI may encourage shallow processing and erode cognitive skills, this study examined whether voice characteristics, conventional (generic) vs. personalized (cloned from the learner), and delivery modes, radiation (via speakers) vs. conduction (via bone-vibrating headphones), shape learners’ ability to reason, not just repeat. Surprisingly, personalized voices enhanced learning under radiation but diminished it under conduction. A second experiment linked these effects to cognitive mindsets: radiation encouraged deeper, analytical thinking, and conduction favored quick, action-oriented processing.

Download

Share

COinS