THANK YOU FOR SUBSCRIBING
In a study conducted by the BMJ, Artificial intelligence technology can bring about plausible, entertaining, and scientifically interesting titles for potential research articles.
FREMONT, CA: A study by the BMJ showcases that the research titles generated through Artificial Intelligence appeared attractive and interesting to the readers. The researchers made use of the BMJ’s 13 most-read Christmas research articles in the past 10 years to prompt similar AI-generated titles to ascertain their study. A sample of 25 doctors from a wide range of specialties in Africa, Australia, and Europe rated the 10 highest and 10 lowest scoring AI-generated titles, which were then clubbed with 10 real Christmas research articles. The results detail how the AI-generated titles are rated, 64 percent of the titles seemed enjoyable, 70 percent attractive and the real titles with 73 percent were rated more plausible. When compared to the real titles, the AI-generated ones were considered less scientific or educational. However, the difference turned out to be negligible when the AI output was curated by humans. Hence, researchers conclude that the finest results are made when there is human interaction with machine learning.
Of all the titles generated,“The clinical effectiveness of lollipops as a treatment for sore throats,” and “The effects of free gourmet coffee on emergency department waiting times: an observational study”, was the most plausible title. “Superglue your nipples together and see if it helps you to stop agonizing about erectile dysfunction at work,” despite the fact that AI fails to decipher the real-world application of a study, this one turned out to be the funniest of all the AI-generated titles. Though they accept that AI-generated titles without human intervention pose several limitations, it is also acknowledged that Artificial Intelligence has the ability to engage and attract readers with their plausible outputs. However, researchers are in constant efforts to highlight the significance of human intervention. They conclude that a finding that mirrors the potential use of AI in clinical medicine, as decision support rather than an outright replacement of clinicians.