2023-08-26 02:25:17
A study carried out by the Harvard Medical School has revealed errors in cancer treatments developed by ChatGPT.
The study, published by the journal JAMA Oncologyfound inappropriate treatment recommendations that were mixed with correct advice.
LOOK: Meet the adorable clouded leopard who is raised at the Oklahoma Zoo
“ChatGPT sometimes speaks in a way that reflects security and seems to make sense. The way in which it mixes correct and incorrect information is potentially dangerous,” explained Danielle Bitterman, an oncologist specializing in the study of artificial intelligence (AI) and its potential application in medicine.
According to a Bloomberg report, which released the study’s findings, the authors were surprised by ChatGPT’s repeated errors when reviewing cancer treatments.
Finally, the specialists highlight the importance of adopting a critical position towards these technologies, and always seek advice from professionals.
RECOMMENDED VIDEOS:
This is how they shot at the car of the journalist who uncovered the Bermejo case
1693030160
#ChatGPT #Recommends #dangerously #flawed #cancer #treatments #artificial #intelligence #health #medicine #JAMA #Oncology #LIFE