AI不釋手：初探台灣民眾對會話型AI的期望確認與持續使用意圖

Enthralled in Chatting with AI: Exploring Taiwanese Users’Expectation Confirmation and Continued Usage Intentions towards Conversational AI

本研究以資訊系統持續使用模型為架構，結合機器直覺思考、感知享樂性、感知似人性等概念，探討台灣使用者在數位語音助理、ChatGPT等兩種會話型AI的使用經驗和持續使用意圖。根據2023台灣網路調查報告資料（N = 749），研究結果顯示正向機器直覺思考能顯著預測使用者對聊天機器人的感知有用性、感知享樂性及感知似人性，而上述三項主觀感知也能預測對機器人的期望確認、滿意度，進而正向影響持續使用意圖。研究發現有助於擴充資訊系統持續使用模型在探討人工智慧產品與服務採用行為時的心理機制與決策過程，可望增進對當代人機互動效果的認識，以期思考人類與AI科技的共存之道。

This study employs the model of information systems (IS) continuance as research framework, integrating concepts such as machine heuristic, perceived enjoyment, and perceived human-likeness, to explore the experiences and continued usage intentions of Taiwanese users with two types of conversational AI: voice assistants and ChatGPT. Drawing from data gathered in the 2023 Taiwan Internet Report (N = 749), the results demonstrated that positive machine heuristic positively predicted perceived usefulness, perceived enjoyment, and perceived human-likeness of both types of conversational AI. Moreover, all three perceptions positively predicted expectation confirmation and satisfaction with the AI applications, thereby influencing users’intentions to continue using them. Research findings expand the aspects of the model of information systems continuance to more comprehensively probe the adoption of AI-based products and services, contributing to our understanding of human-machine interactions and stimulate contemplation on the synergy between human beings and AI technologies.