Research published in the journal Judgment and Decision Making examined how people perceive and evaluate short stories created by artificial intelligence compared to those written by humans. The study involved 1,682 adults who each read one of six short stories—three authored by humans and three generated by ChatGPT, with each AI piece sharing thematic elements with a human-written counterpart.
Participants were informed whether their assigned story came from a human or AI source, though this information was sometimes deliberately inaccurate. Researchers then assessed how the readers rated the stories’ absorbing qualities, engagement level, and overall quality. Results indicated that those who read AI-generated stories rated them as more absorbing and of superior quality compared to participants who read human-authored stories. Conversely, participants assigned higher ratings to stories they believed were written by humans, suggesting that attribution influences perception.
Dr. Deena Skolnick Weisberg from Villanova University noted that AI systems have demonstrated the capacity to generate short stories comparable to or surpassing human work. She emphasized, however, that this capability does not necessitate eliminating human creativity from the literary landscape, as writing serves important functions including self-expression and personal meaning-making that extend beyond mere content generation.
Additional experiments involving 905 participants tested whether readers could distinguish human-written from AI-generated stories when presented with paired examples on identical themes. Performance was inconsistent across trials, with accuracy rates ranging from approximately 40 percent to 52 percent—results that suggest people struggle to reliably identify the authorship source. Participants with greater familiarity with AI systems performed somewhat better at this discrimination task than those with less exposure.
Weisberg attributed the preference for human-authored stories to an authenticity bias, while suggesting that AI stories may actually appeal more due to their accessibility and readability. The research highlights a disconnect between actual and perceived quality, as well as the influence of consumer attitudes toward artificial intelligence on their evaluations.
Article Attribution | Read More at Article Source
Article summary produced by Claude AI