ARTIFICIAL Intelligence bot ChatGPT has a better bedside manner than real doctors, a study shows.
It was judged to give higher quality and more empathetic answers to real-world health questions 79 per cent of the time.
Dr Aaron Goodman, of University of California San Diego, said: “I never imagined saying this but ChatGPT is a prescription I’d like to give to my inbox.
“The tool will transform the way I support my patients.”
ChatGPT was launched last year by OpenAI and uses one of the most sophisticated language models ever developed.
Previous research showed it was able to pass the US Medical Licensing Exam, a test taken by students before they can practise.
The latest study, published in JAMA Internal Medicine, examined if the bot could respond accurately to questions patients sent to doctors online.
Researchers compared 195 responses from doctors with answers given by the bot.
Three licensed healthcare professionals were asked to assess the answers without knowing who gave them — and 79 per cent of their preferred responses came from the chatbot.
But Prof Anthony Cohn, of the University of Leeds, said: “Given ChatGPT’s propensity to invent ‘facts’ and hallucinate, it would be dangerous to rely on any factual information given.”
