AI-Written Stories Rated Higher Than Human Works

A new study has found that readers often rate short stories generated by artificial intelligence as more engaging and of higher quality than comparable human-written works, while also struggling to distinguish between the two. The research, published in the journal Judgment and Decision Making, suggests public perceptions of AI's creative abilities may be lagging behind its technical performance.

Conducted by researchers at Villanova University, the study involved 1,682 U.S. adults who each read one of six short stories, all roughly 1,000 words of realistic fiction. Three stories were written by published human authors, while three were generated by ChatGPT 4.0 with similar themes. Participants were told whether the story was written by a human or AI, but these labels were sometimes intentionally incorrect. After reading, they rated the stories on quality and how absorbing they found them.

The results showed that AI-generated stories received higher average ratings than the human-written ones. However, the assigned label significantly influenced opinions. Stories participants believed were written by humans received higher scores, even when those stories were actually...