This summary is provided by the IPR Digital Media Research Center.
Rachel Son and colleagues analyzed whether audiences were as satisfied with machine-generated script narratives (bot narratives) as they were with scripts written by humans.
An online survey of 260 U.S. adults was conducted in May 2023.
Key findings include:
1.) Participants’ levels of narrative understanding, focus, and emotional engagement did not vary when reading the human script versus the bot authored script.
2.) Readers enjoyed the stories more when they thought they were written by humans.
— Participants responded differently to the stories when they were told the author was AI-generated as opposed to human-generated, even though the content was identical.
3.) Narrative elements are less influential on people’s perception of quality and their level of enjoyment than author labels.
— The perceived identity of the author influenced reader enjoyment and perception of quality more than story content.
4.) Human author labels failed to enhance participants’ perceived authenticity of the scripts.
— Scripts that were labeled as human-authored, regardless of whether they were generated by robots, resulted in participants perceiving higher levels of authenticity compared to scripts labeled as bot-authored.
Find the original research here.
