The AI model ChatGPT has in a study from Uppsala Academic Hospital shown itself to be ten times faster than real doctors in performing administrative tasks, reports SVT Uppsala.
In the study, doctors and ChatGPT performed the same administrative tasks regarding six fictional patient cases, which were then assessed by an expert panel.
The quality between the doctors' and the AI-generated tasks could not be distinguished. However, ChatGPT was much faster in completing the job.
Cyrus Brodén, orthopedic surgeon and researcher at Uppsala Academic Hospital and Uppsala University, believes that ChatGPT in the future can facilitate administrative work in healthcare, but urges caution.
One is positive, but everyone agrees that it is important to really evaluate the tools before they are introduced into healthcare, so there is a healthy skepticism, he says to SVT Uppsala.