台灣民眾使用生成式AI意願之探索

A Study on Taiwanese People’s Willingness to Use Generative AI

生成式人工智慧（GAI）科技的成熟應用，帶給人類一個知識悖論課題：GAI的應用已可執行過去人類專屬的任務（例如自然的生活對話或流暢的文章寫作），且表現非常的出色。然而從傳統資訊計算的研發品質上（例如可靠性、準確性和真實性），卻很難符合統計科學驗證。因此，本研究以實務應用面角度，探討台灣民眾對於生成式人工智慧應用程式的知覺優勢、劣勢、效益與風險，及其和計畫行為理論的科技態度、主觀規範、行控知覺與行為意願等基本變數間之關聯性。本文採量化研究法，通過匿名填答的線上問卷收集來自台灣地區277位民眾數據，並以PLS-SEM統計法進行數據分析。結果顯示，台灣民眾知覺到的GAI工具優勢、劣勢、效益和風險分別對科技態度、主觀規範和行控知覺有不同的顯著正向影響。其次，TPB理論的三個自變項對於GAI應用程式的使用意圖有正向顯著影響，而此種意圖對於他們實際應用GAI工具行為亦產生正向影響。本文通過理論彙整出可能影響台灣民眾使用GAI工具意願的相關因素，希冀能夠提供實務洞悉與見解，讓台灣GAI產業利害關係人得以制定合適的產業發展規則和政策，且為台灣地區相關文獻進行補充。

The rapid advancement and maturity of generative artificial intelligence (GAI) technologies have introduced a cognitive paradox for humanity. GAI tools are now capable of performing tasks that were once considered uniquely human-- such as engaging in naturalistic everyday conversations or composing coherent and articulate written content-- with remarkable proficiency. However, from the perspective of traditional information computing standards, including reliability, accuracy, and verifiability, GAI tools often fall short of meeting the criteria required by statistical scientific validation. Accordingly, this study adopts a practical application perspective to investigate Taiwanese citizens’perceptions of the advantages, disadvantages, benefits, and risks associated with generative AI applications, as well as the relationships between these perceptions and the core constructs of TPB, namely attitude, subjective norms, perceived behavioral control, and behavioral intention. A quantitative research approach was employed, using an anonymous online survey to collect responses from 277 individuals in Taiwan. The data were analyzed through PLS-SEM. The results indicate that perceived advantages, disadvantages, benefits, and risks of GAI tools each exert varying degrees of significant positive influence on attitudes, subjective norms, and perceived behavioral control. Furthermore, all three TPB antecedents demonstrated significant positive effects on users’behavioral intention to adopt GAI applications, which in turn positively influenced their actual usage behavior. This study synthesizes theoretical insights to identify the key factors affecting Taiwanese users' willingness to adopt GAI tools. It aims to provide practical implications and actionable insights for stakeholders in Taiwan's generative AI sector, enabling them to formulate appropriate industry development guidelines and policy measures. Additionally, the findings contribute to enriching the body of literature related to generative AI usage in the Taiwanese context.