Abstract
Artificial intelligence-enabled (AI-enabled) conversational agents have gained global popularity due to the development of artificial intelligence capabilities and the benefits to human lives and society. However, research on the adoption of AI-enabled conversational agents is in an early stage, and there is a lack of understanding of why people adopt AI- enabled conversational agents. To explain this underexplored phenomenon, this thesis aims to explore the factors that influence the adoption of AI-enabled conversational agents.This thesis consists of three studies, each presenting a different perspective. The first study explains how different types of perceived value (i.e., utilitarian value, hedonic value, novelty value, friend value, and master value) shape the adoption of AI-enabled conversational agents. Specifically, the literature on perceived value is employed to identify utilitarian value, hedonic value, and novelty value, which are the most common value people perceived in prior research that positively influences adoption intention. Besides, drawing on the computers are social actors paradigm, friend value and master value are proposed to be another key driver of adoption intention. Empirical validation is carried out through an online questionnaire. The results confirm the key roles of friend value, master value, utilitarian value, and hedonic value, but not of novelty value, in the intention to adopt. The findings suggest the dual nature of AI-enabled conversational agents as both tools and social entities. To fully understand the adoption of AI-enabled conversational agents, research should cover the perspective of human-AI relationships.
The second study explains how the memory of experiences contributes to the adoption of AI-enabled conversational agents. Specifically, drawing on the technology acceptance model and theory of planned behavior, this study focuses on individuals with some generative artificial intelligence (AI) experiences to understand technology adoption. This study proposes that the memory of experiences is a significant antecedent of present beliefs, present attitudes, and adoption intention of generative AI. The research model is validated through a survey study. Key findings highlight the significant effects of the memory of experiences on present beliefs, present attitudes, and adoption intention. The results reveal the value of the memory of experiences in information systems research, and taking the perspective of the memory of experiences can provide insights into the phenomena associated with individuals and technology.
The third study develops and tests a research model that examines the complex nature of the viral adoption phenomenon in the context of AI-enabled conversational agents. The third study introduces the concept of adoption speed to capture the temporal aspect of adoption and draws on both technology adoption and sharing literature to propose adoption potential and sharing as important antecedents of adoption speed. Leveraging concepts from disruptive innovation and sharing, this study theorizes that people’s beliefs regarding the adoption of ChatGPT and its technological characteristics play a pivotal role in shaping adoption potential and sharing. An online survey is employed to validate the proposed model. The results confirm that adoption potential and sharing positively influence adoption speed, potentially leading to viral adoption. Results also show that people may not necessarily share their own personal benefits derived from technology. Viral adoption captures a new facet of the technology adoption phenomenon, which warrants attention from IT companies and scientific communities.
This thesis employs three different angles to understand the adoption of AI-enabled conversational agents, focusing on the human-AI relationship, the memory of experiences, and viral nature of adoption. The thesis delves into not only why people adopt AI-enabled conversational agents massively (in the first and second studies) but also why people adopt AI-enabled conversational agents virally (in the third study). From the theoretical perspective, this thesis contributes to technology adoption research by introducing three new perspectives and serving as a foundation for future studies. This thesis also provides practical insights into AI technology’s design and marketing strategies.
| Date of Award | 28 Aug 2024 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Christian WAGNER (Supervisor) & Ayoung SUH (Supervisor) |
Cite this
- Standard