BERLIN: In the latest worrying sign that people are increasingly incapable of distinguishing artificial intelligence (AI) from themselves, Cambridge University Press (CUP) has published research showing that readers give top billing to chatbot-generated writing after being misinformed that items were written by a human.
The near-2,000 participants in the study were "unable to tell the difference between human- and AI-created stories,” according to researchers at Villanova University, whose paper was published in the CUP journal Judgment and Decision Making.
"People gave the highest ratings to stories generated by AI that they were told had been written by a human,” CUP says in a statement launching the findings. These add "to the growing body of evidence that many people’s ability to spot AI-generated content is limited,” the publisher says.
Only four in 10 of the participants could correctly identify which short stories were churned out by a chatbot, the Villanova team found.
Attempting to explain why so many of those surveyed thought the AI content they were reading was written by a person, Deena Weisberg, a psychologist at Villanova, says the outcome shows "public assumptions” about AI to be "increasingly out of date.”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,” Weisberg says.
A survey published in June and carried out by Internet security platform Malwarebytes showed nine out of 10 people were unable to tell the difference between real and fake on the web, with AI blamed for the confusion.
And in July, a team of German researchers found AI impersonations of political debaters came across as more "authentic, relevant and coherent” than the real thing.
In further concerning research, a team in China said in March that even after training, people are increasingly struggling to tell the difference between real voices and AI-generated mimicry. – dpa
