Artificial Intelligence outclassed a Physician doctor in Intelligence Quotient Test

 

Artificial intelligence: image by Freepik

AI Outperforms Human Doctors in Medical Quiz, Demonstrating Empathy and Knowledge

In a groundbreaking study conducted by the University of California, the question of whether an AI could provide knowledgeable and comforting responses to medical queries, similar to a human doctor, was explored. The study compared the performance of physicians and the AI chatbot ChatGPT in answering 195 medical questions, focusing on their empathetic capabilities.

To gather the questions, researchers utilized the 'AskDocs' subreddit, where individuals sought advice on various medical concerns. Examples included inquiries about the dangers of swallowing a toothpick or the likelihood of a concussion after hitting one's head on a metal bar.

A verified healthcare professional responded to the questions on the subreddit, while ChatGPT generated its own answers. To determine the superior response, a panel of physicians blindly evaluated each case without knowing whether it was the chatbot or the human doctor who provided the answer. The panel consisted of three judges, and their evaluations were averaged, resulting in a total of 585 assessments.