Actors, Agents, and Avatars: Visualizing Digital Humans in E-Commerce and Social Media
Permanent URI for this collectionhttps://hdl.handle.net/10125/107504
Browse
Recent Submissions
Item type: Item , Who Sells Better? Digital Human Presenter Versus Cartoon AI Presenter in E-commerce Live-Streaming(2024-01-03) Ma, Hanzhuo; Huang, Wei; Dennis, AlanAI technology has been introduced on e-commerce live-streaming as a substitute for human presenters who are expensive and constrained by work time. However, the cartoon avatar AI presenters that are commonly used today are less effective in driving sales. We propose that increasing human likeness is a practical method to boost sales, particularly for products with hedonic values and rich sensory attributes. The results show that a digital human presenter (an AI agent with a highly human realistic face and voice) sold better compared to a cartoon avatar presenter across a range of different types of products and that this effect is driven both by affect (by increasing positive emotions) and cognition (by improving product quality evaluations). We extend current research on AI agent design to the live-streaming context and show that the highly realistic appearance of digital humans, as a powerful design, can enhance positive consumer responses to AI agents through two parallel theoretical routes (emotional and cognitive). These results can help e-commerce platforms and brands improve their AI technology and obtain expected business benefits.Item type: Item , Using Chatbots and Digital Humans to Collect Online Reviews(2024-01-03) Rosengren, Warren; Sachdeva, Agrim; Kim, Antino; Dennis, AlanAI-controlled digital humans that look and sound highly human-realistic are beginning to appear in place of text-based chatbots for a variety of routine customer service tasks. We used an online experiment to examine how people respond to digital humans and chatbots when providing online reviews. The results show that participants perceived the digital human interface to be more effective, efficient, and satisfying. Participants found the digital human to be more human-like, which elicited a stronger emotional response and led to higher satisfaction. Participants used more casual and friendly language when interacting with the digital human and provided longer feedback than with the chatbot. Overall, people preferred the digital human interface over the chatbot. These findings suggest that using highly realistic digital humans in customer service could be a beneficial and worthwhile option to consider.Item type: Item , Can We Trust an AI Agent? Interaction Effects of Its Machine Learning Performance and Digital Character(2024-01-03) Kim, Tae Jin; Lee, One-Ki Daniel; Kang, JuyoungAI-powered digital characters, AI agents, are expanding their scope of application to various fields. However, research on the key factors influencing consumer attitude is insufficient. This experimental study focuses on machine learning (ML) performance (i.e., the behavioral (intelligence) realism of AI agents), which determines users’ trust. This study further investigates the interaction role of the different forms of digital character (i.e., the form realism of AI agents) in the relationship between ML performance and trust. The findings of this study provide a novel understanding of human-AI interaction, expand academic understanding of AI anthropomorphism, and suggest new research directions for digital humans. The results will also guide business practitioners in developing AI services.Item type: Item , Introduction to the Minitrack on Actors, Agents, and Avatars: Visualizing Digital Humans in E-Commerce and Social Media(2024-01-03) Riemer, Kai; Seymour, Mike; Yuan, Lingyao