STUDYING HOW INTERACTION WITH ARTIFICIAL INTELLIGENCE CAN HELP SHAPE HUMAN BELIEFS AND INFLUENCE ATTITUDE CHANGE: EXPLORING GLOBALIZATION PERSPECTIVES (2023-24)
Hanqiu Li
Ph.D. student in management and organizations (entered 2023)
Artificial intelligence (AI) has become a widespread presence in many technological products that we use on a daily basis. OpenAI’s ChatGPT (AI chatbot) has been a notable contributor to this. The increasing use of ChatGPT has led to a phenomenon known as anthropomorphism, where users attribute human-like qualities to AI tools. This process, as outlined by Epley et al. (2007), involves projecting human characteristics onto non-human entities, transforming interactions from human-to-object to human-to-human-like. This shift can foster a sense of attachment as the AI begins to fulfill emotional and psychological needs related to comfort, self-identity, and self-efficacy. This anthropomorphism and resulting attachment are important factors in developing trust in AI, which much like trust in human interactions, can influence beliefs. The ways in which different countries use AI and the resulting implications of its use in shaping people’s ideology and behaviors are creating new global cooperation opportunities in solving global societal issues.
Hanqiu Li’s research seeks to understand the potential of AI in shaping public opinion and motivating action on critical global societal issues. At the heart of Li’s research is the question: Does the attitude that one is conversing with AI (e.g., the use of ChatGPT) rather than a human lead to a more significant change in viewpoints on certain global issues, including political, societal and economic perspectives? Her study explores whether interacting with AI, particularly through trust in AI, can shape users' beliefs and ideologies, similar to interactions with trusted humans. Additionally, the study seeks to assess if such belief changes translate into a willingness to take action related to these global challenges.
Financial support from the CGM was critical to provide the resources needed to develop and implement the experimental setup that focused on three conditions: perceived AI and behavioral AI, perceived AI and behavioral human, and perceived human and behavioral human. Funding was used to recruit a coder who wrote the code necessary to build the chatbot for the study, which enabled the creation of a robust and interactive AI system for the experiment. Additional funding was used for participant recruitment for baseline data collection to ensure a diverse and adequate sample size for reliable data analysis. Overall, 15 pilot interactions were conducted and data from over 700 participants was collected over the 14-day study period.
The study generated rich datasets encompassing participants' interactions with AI, their responses to surveys, and behavioral changes over the study period. It also included metrics on belief changes in the three areas of focus and the participants’ willingness to take action in these areas. Detailed logs of interactions between participants and the AI were recorded, providing insights into how AI-mediated communication influences belief and behavior change. The next phase of the project will be to analyze the data and results from the experiment to measure and evaluate any shifts in beliefs and attitudes toward the three perspectives of globalization (political, societal and economic) and to generate a preliminary set of findings. The findings could further provide scientific implications of whether governments should get involved in the process of AI development by working with AI companies to set up standards and regulations.
Findings from this project will be of great business and management relevance. Li’s research addresses real-world business, economic and political decisions, and will contribute to the goal of advancing global management thinking and societal relevance. The outcomes of her study will have potential implications for policy development and ethical standards in AI usage, and offer insights into the evolving role of technology in shaping public opinion on important global societal issues. The study will also offer insights into how global leaders can leverage AI to communicate and manage cross-culturally, respecting and understanding diverse perspectives. Understanding how AI can influence beliefs and actions toward global issues can also inform public policy and social initiatives. The interdisciplinary nature of the research, bridging AI, psychology, ethics and management, exemplifies the CGM’s goal to foster cross-disciplinary collaboration in addressing complex global issues. This research will be presented at academic conferences, fostering further scholarly discussion and collaboration. Li will also be submitting the findings to various publications and journals in the fields of management, AI and social sciences.