Paper Type
Complete
Paper Number
PACIS2025-1258
Description
Human-AI interaction has undergone a significant transformation with the advancement of artificial intelligence, particularly marked by the shift from rule-based systems to sophisticated generative AI models. In recent years, the rise of generative AI and large language models has revolutionized these interactions by enabling machines to generate contextually relevant and coherent human-like texts. Voice-enabled agents, powered by text-to-speech technology, produce human-like speech by converting written text into spoken words. This study draws on the acoustic literature to explore two critical aspects of voice synthesis: voice production and voice transmission. By leveraging the dual-stream model of auditory processing, we propose a research model to investigate human-AI interactions in experiential learning contexts. We conducted two laboratory experiments within a controlled human-AI interaction environment to test our hypotheses. Ultimately, this study contributes to the field of human-AI interaction by revealing the complex relationship between information processing strategies and inherent human biases toward AI.
Recommended Citation
Zhu, Fenfen; Choi, Ben; and Tishchenko, Viktoriia, "Hello Jo, Help Me Please: The Impact of Synthetic Voice on Human Psychology and Performance" (2025). PACIS 2025 Proceedings. 3.
https://aisel.aisnet.org/pacis2025/aiandml/aiandml/3
Hello Jo, Help Me Please: The Impact of Synthetic Voice on Human Psychology and Performance
Human-AI interaction has undergone a significant transformation with the advancement of artificial intelligence, particularly marked by the shift from rule-based systems to sophisticated generative AI models. In recent years, the rise of generative AI and large language models has revolutionized these interactions by enabling machines to generate contextually relevant and coherent human-like texts. Voice-enabled agents, powered by text-to-speech technology, produce human-like speech by converting written text into spoken words. This study draws on the acoustic literature to explore two critical aspects of voice synthesis: voice production and voice transmission. By leveraging the dual-stream model of auditory processing, we propose a research model to investigate human-AI interactions in experiential learning contexts. We conducted two laboratory experiments within a controlled human-AI interaction environment to test our hypotheses. Ultimately, this study contributes to the field of human-AI interaction by revealing the complex relationship between information processing strategies and inherent human biases toward AI.
Comments
AI ML