In the literary industry, using artificial intelligence is still a taboo. Even accusations of AI use have dire consequences, from the infamously canceled release of the novel Shy Girl to the massive online backlash against a prestigious literary award going to an apparently AI-written short story.
Like all AI-generated art, creative writing crafted by a bot seems to be missing something fundamentally human. But as AI-generated content advances and becomes harder and harder to distinguish from human-made work, is people’s distaste for robot writing based on the product itself, or merely on how it was made?
A new study suggests the latter: Though readers give preference to stories they think are written by humans, ultimately, they like AI-written pieces the best.
A pro-human bias
The study, published by Cambridge University Press, asked 1,682 adult participants to read one of six short stories. Three of the stories were written by humans. The other three were generated by ChatGPT, each with a similar topic and style to one of the human-written samples.
Each reader was told whether the story they read was written by a human or by AI—but they weren’t always told the truth.
Overall, readers rated AI-written stories as more engaging and higher-quality than their human-written counterparts. When participants were told the AI stories were written by humans, they gave them even higher scores, suggesting an implicit bias against AI-generated content regardless of its quality.
One of the study’s authors, Deena Skolnick Weisberg of Villanova University’s Department of Psychological and Brain Sciences, said in a press release that “public assumptions about AI’s capabilities are increasingly out of date.”
“We assume creative writing requires uniquely human qualities, such as emotional understanding and lived experience, which leads people to underestimate AI’s capabilities,” she said.
What sets AI writing apart?
AI-generated writing is trained on human work, and this experiment specifically called for AI stories based on existing human-written narratives. What, then, makes AI writing more appealing to readers?
Weisberg thinks it may have to do with AI writing’s tendency to be “clearer, more direct, and easier to process,” she said. “By contrast, human-written stories are often more subtle and complex.”
“Perhaps people generally prefer predictability, because difficult or subtle material requires more brain power,” she added. AI writing certainly fits the bill of predictability: A recent report identified current trends in AI-generated content, including infrequent punctuation, long sentences, and a reliance on rhetorical conventions like “it’s not X, it’s Y” and the rule of threes.
A different study published earlier this year found that large language model (LLM) usage contributed to the “blandification” of human writing, producing copy that authors deemed less creative, less passionate, and more homogenous in tone than unassisted human writing. Despite this, authors who used LLMs were just as satisfied with their output as those who didn’t, indicating that humans are becoming more and more comfortable with AI-written content.
But even Weisberg reasons that artificial intelligence can never truly stand in for human writers.
“We may need to make room for AI-generated novels, and for AI/human co-written novels, but that doesn’t mean that there’s no longer space for us to appreciate the process of human creativity,” she said. “Humans write as a form of creative self-expression or to challenge ourselves, or to make sense of our experiences. The fact that AI can generate human-like stories doesn’t change any of that.”
