Artificial Politics: The Role of Artificial Intelligence in Shaping Political Action
Format
Oral Presentation (Research Day, April 30)
Faculty Mentor Name
Dari Sylvester Tran
Faculty Mentor Department
Political Science
Abstract/Artist Statement
If the 2016 Presidential elections in the United States served as an example for anything, it is that for better or for worse AI is changing the political landscape. Artificial intelligence (AI) is becoming ubiquitous across disciplines, from medical AI capable of diagnosing breast cancer to text-based AI capable of writing articles about sports, science, and politics. In this research study, we seek to develop and extend our current understanding of AI’s effects on political belief formation by focusing on AI-generated news articles and tweets. We have designed two methods for this study: (1) a statistical analysis examining the relationship between AI robots on Twitter and people’s political beliefs on Brexit and (2) a survey-based experiment focusing on isolating a relationship between a person’s political beliefs and exposure to news articles written by AI. In the first experiment, we found that there is a positive relationship between exposure to AI content about an issue online and human users’ beliefs about that issue. The second experiment also indicates that there is a small, positive relationship between exposure to AI-generated political content and changes in an individual’s political opinions about the content’s subject matter.
Location
Yosemite Learning Lab, William Knox Holt Memorial Library and Learning Center
Start Date
30-4-2022 2:00 PM
End Date
30-4-2022 2:19 PM
Artificial Politics: The Role of Artificial Intelligence in Shaping Political Action
Yosemite Learning Lab, William Knox Holt Memorial Library and Learning Center
If the 2016 Presidential elections in the United States served as an example for anything, it is that for better or for worse AI is changing the political landscape. Artificial intelligence (AI) is becoming ubiquitous across disciplines, from medical AI capable of diagnosing breast cancer to text-based AI capable of writing articles about sports, science, and politics. In this research study, we seek to develop and extend our current understanding of AI’s effects on political belief formation by focusing on AI-generated news articles and tweets. We have designed two methods for this study: (1) a statistical analysis examining the relationship between AI robots on Twitter and people’s political beliefs on Brexit and (2) a survey-based experiment focusing on isolating a relationship between a person’s political beliefs and exposure to news articles written by AI. In the first experiment, we found that there is a positive relationship between exposure to AI content about an issue online and human users’ beliefs about that issue. The second experiment also indicates that there is a small, positive relationship between exposure to AI-generated political content and changes in an individual’s political opinions about the content’s subject matter.