初探台灣民眾對人工智慧產品與服務之採用經驗與信任感

Abstract

近年來人工智慧逐漸出現在台灣民眾的日常生活中，而且產品與服務日益多元。本研究結合創新傳布與調節焦點理論，比較不同使用者特質的民眾（先驅者、猶豫者、跟風者及觀望者）對AI科技所具備相對優勢的觀感差異，進而對整體採用經驗的影響。同時，也將四項AI產品與服務的擬人化程度，再區分為低擬人型（智慧手環／手表、AI金融軟體）和高擬人型（數位語音助理、聊天機器人），以媒介等同論為基礎，進一步檢視民眾採用此兩類型AI科技的經驗與其AI信任感之間的關聯性。研究以2022年台灣網路調查報告資料（N=1,725）為數據，結果顯示，先驅者最看好AI科技相比傳統人力服務所具有的相對優勢，而民眾愈看好AI所具備的相對優勢，整體AI產品與服務也採用愈多。不過，民眾愈有採用高擬人型AI的經驗，對AI的信任感愈高，但是對低擬人型的AI愈具採用經驗，則與信任感無顯著相關。研究結果一方面可擴充上述經典理論在新興科技採用議題上的適用性，另一方面亦有助於AI科技在市場推廣及相關政策制定時參考。

This study integrates diffusion of innovations, regulatory focus theory, the media equation, and the tenets of anthropomorphism as research framework to understand Taiwanese users' adoption of artificial intelligence (AI) technologies. In addition, the relationships between user adoption of lower (smart wearables and AI financial software) and higher (digital assistants and chatbots) anthropomorphic AI technologies and their level of AI trust are investigated. Drawing upon the 2022 Taiwan Internet Survey database (N = 1,725), the results indicated that the pioneer"-compared to the "hesitator," "bandwagoner," and "fence-sitter"-reported the highest level of perceived AI relative advantage, which in turn promoted their adoption of both the lower and higher anthropomorphic AI technologies. Yet, only adopting higher anthropomorphic AI services and products demonstrated a significantly positive effect on AI trust. The findings help elucidate the resemblance between interpersonal communication and human-machine interaction, drawing implications to marketers and policy makers in promoting the use of AI technologies."