A new study has found that audiences perceive human writers to be more authentic than machine authors. The findings by University of Florida College of Journalism and Communications (UFCJC) doctoral students Rachel Son, Qingyuan Yang, and Benjamin Vollmer and Advertising Associate Professor and STEM Translational Communication Center Interim Director Benjamin Johnson,  were featured in “Engaging with Bot Narratives: The Audience Experience and Perception of Machine Script Writers.” The paper was presented at the 73^rd annual International Communication Association Conference in Toronto, Canada in May 2023.

In addition, Son and Johnson collaborated with UFCJC doctoral student Chelsea Moss on the book chapter “Exposure to and Effects of International Entertainment on Streaming Platforms: A Review and Agenda for Future Media Psychology Research” to be published in Television’s Streaming Wars in November 2023.

For the study of bot narratives, the authors conducted an experiment that tested how readers respond to a short piece of fiction, depending on whether the story has stylistic cues of being bot versus human-written and on whether the story is explicitly labeled as bot versus human-written. They examined key outcomes widely used in entertainment research, including narrative engagement and enjoyment, as well as perceptions of the author and their intelligence.

According to the authors, “Results show that the author of the story did not significantly predict readers’ level of narrative engagement, enjoyment, or perception of the author’s authenticity, humanness, or mind. Whether participants were reading the human or machine authored story, individuals did not vary in their level of narrative understanding, attentional focus, narrative presence, or emotional engagement. However, the label assigned to the author did predict the participants level of narrative engagement and their perceptions about the author to some extent.”

They add, “Whether it is the machine author, or the author only being labeled as such, audiences perceive human writers to be more authentic. With AI technology, such as ChatGPT, still in its infancy, future research is needed to continue to understand how different machine capabilities and characteristics influence the audiences’ enjoyment and experience with machine-generated entertainment content.”

Son, Moss and Johnson’s chapter on streaming platforms examines potential effects of streaming international entertainment, as well as the potential for dynamic effects over time.

According to the authors, “Viewers may be less likely to select foreign over domestic narratives, yet a variety of motivational factors, message factors, and situational factors can contribute to interest in international entertainment. Next, the entertainment experience entails information processing, including character, narrative, and emotional engagement. We examine available evidence and theory about how international entertainment is processed by viewers. After processing, these narratives can yield affective, behavioral, and cognitive media effects.”

Posted: July 5, 2023
Category: AI at CJC News, College News, Student News
Tagged as: AI at CJC News, Benjamin Johnson, Benjamin Vollmer, Chelsea Moss, Rachel Son