Study: Readers Rate AI Stories Higher Than Human Ones
Get the Health newsletter
Daily health & science — research, biotech, public health, the studies worth knowing. Free.
- Judgment and Decision Making published a study in which 1,682 adults read one of six short stories — three written by humans and three by ChatGPT — with each AI story matched thematically to a human-authored counterpart.
- Participants who read an AI-generated story rated it as more absorbing and of higher quality than those who read a human-written one, yet simultaneously gave higher ratings to stories they had been told were human-authored.
- Dr Deena Skolnick Weisberg of Villanova University, a senior author and creative writer herself, said AI can already produce short stories seen as at least as good as human work but stressed that human creative expression still has its own distinct value.
- Two follow-up experiments with 905 adults asked to identify which of two stories was human-written produced correct-guess rates of roughly 40% (worse than chance) and 52%, leading the authors to conclude participants could not generally tell AI and human stories apart.
- Greater AI expertise, not fiction-reading expertise, correlated with better ability to identify AI-written stories, with Weisberg comparing the process to learning to recognise an individual author's style.
- Luke Kennard, a professor of film and creative writing at the University of Birmingham, rejected the framing of AI as a neutral tool, calling it "predictive code based on a massive stolen database" and an "existential threat" and arguing the technology's environmental and social costs make it categorically not worth using.
Why it matters: With 1,682 participants rating AI stories more highly than human-written ones — and follow-up experiments showing readers essentially could not tell them apart — the study adds empirical weight to a debate already reshaping creative industries. For publishers, educators and writers' groups, the finding raises practical pressure on how AI-authored work is disclosed, labeled and compensated, while the environmental and ethical critique raised by Kennard keeps the human-versus-machine framing unsettled.
