时间：2026年7月9日 9:45-10:30

地点：东校区财科馆317

题目：Text-to-Image AI and the Novelty of User Generated Content on Social Media

摘要Abstract：

Text-to-image AI tools enable social media users to transform their imaginative ideas into visual content with different visual styles, adding a new type of content for social media platforms. We argue that the AI tools pose a spillover effect on the novelty of user-generated regular content on social media (i.e., content generated without an AI tool) potentially through two drivers: a usage-driven fixation mechanism rooted in the process of interacting with AI tools, and a reputation-driven mechanism related to social feedback systems. Collaborating with an image-sharing social media platform, we employ a difference-in-differences (DID) approach to empirically demonstrate that using a text-to-image AI tool significantly reduces the novelty of user-generated regular images. That is, after using the text-to-image AI tool, the stylistic features of AI users’regular images become more similar to recent platform-wide images than those of non-AI users. Also, empirical evidence consistently validates the existence of an AI-usage fixation mechanism. Novelty reduction worsens with more frequent AI usage, and regular images from AI users become more similar to the AI-generated images than those from non-AI users. The conformity to AI suggests that frequent AI use may induce a fixation effect. Delving into more granular image-level generation records, we conduct an instrumental variable analysis and find that repeatedly generating AI images in a specific style leads to greater conformity of their regular images to that AI style, implying that the repeated use of the same AI style intensifies the fixation effect. Finally, AI users who consistently input prompts with a narrower informational scope and less semantic variation (i.e., those who are more subject to the fixation effect) exhibit greater novelty reduction in regular content. However, we do not find evidence supporting the reputation mechanism and other explanations.

主讲人简介：Guangzhi Shang is an Associate Professor in the Department of Supply Chain Management at W.P. Carey School of Business, Arizona State University. Guangzhi’s current research has three primary themes: consumer returns management, service labor issues, and management of innovative technologies. He investigates the first from a variety of angles, including how a retailer should set its optimal return policy, how an OEM or a retailer could better forecast the quantity of returns, and how a retailer could assess the value of its return policy. For the second, he focuses on the context of live-chat contact centers. Research questions include the impact of customer’s waiting experience on the progress of a chat session, agent’s ability to learn from their past experiences, and the customer–agent matching problem. For the third, he looks into emerging fintech such as cryptocurrency and crowdfunding platforms. He enjoys doing practice-driven research. He is a frequent invited speaker at leading industry conferences such as the annual Consumer Returns conference. He is among the first to publish on the topic of cryptocurrency and blockchain in premier business school journals.

His research has been published in Manufacturing & Service Operations Management (MSOM), Production and Operations Management (POM), Journal of Operations Management (JOM), MIS Quarterly (MISQ), and Decision Sciences (DS), among others, and recognized by best paper awards at POM, JOM, and POM Society’s College of Operational Excellence. He serves as the Department Editor for the Empirical Research Methods Department at JOM and for the Retail Operations Department at DS. His review service is recognized by the 2019 outstanding reviewer award of DS and the 2018 best reviewer award of JOM. He was also nominated for the best reviewer for POM and best associate editor for JOM. He co-produced a column together with Mike Galbreth from University of Tennessee and Mark Ferguson from University of South Carolina in the Reverse Logistics Magazine named “View from Academia,” aimed at disseminating fresh-off-the-press academic knowledge among industry professionals dealing with consumer returns. He currently co-edits a special issue for JOM: “Operational Perspectives on Blockchain Applications”. He has been interviewed by Fox News and WCTV for his expertise on cryptocurrency and blockchain.