According to a Mass Common Brigham news release, the study, published in the Journal of Medical Web Analysis, revealed that ChatGPT was around 72% correct when it came to common decision-making from the first point of contact with a patient.