ChatGPT has passed the medical specialty test Do doctors have artificial intelligence in the pipeline?

ChatGPT was able to pass the tough exam necessary to become a doctor at the level of the USA. Researchers assessed the accomplishment of the chatbot's artificial intelligence as a significant milestone in the field of medical science.

Chatbot ChatGPT passed between 52.4 percent and 75 percent during different sections of the 3-part Medical Licensing Examination (USMLE) in the USA. To successfully pass the exam it is required to be able to pass the exam with an average score of 60 %.


Researchers from the company AnsibleHealth that carried out the study, stated in a statement about the subjectthat “Reaching the required score for the notoriously difficult expert test and doing it with no human intervention is an extraordinary achievement in the field of artificial intelligence in clinical settings.” The researchers used the words.


The results of the study are published in PLOS Digital Health. It is which is a peer-reviewed journal of science. The research was not without controversy, however, ChatGPT had previously been evaluated and approved in the departments of business (University of Pennsylvania’s Wharton School of Business) as well as department of law (University of Minnesota) departments.

In the most recent study, researchers evaluated the program with 350 questions from the USMLE exam in June 2022. Two physicians reviewed the findings and any discrepancies were analyzed by a third doctor. The test tests the medical expertise of doctors and medical students in training. It has been administered since 1992.

The medical students of the US typically pass part in step one on the USMLE at the conclusion the second semester at medical school. Step two is in their fourth year and step three during their initial year in the residency following having completed medical school. More than 100 thousands of undergraduate and postgraduate students sit for the test each year.

You may like :   Microsoft Starts Restricting Emulator Use on Xbox Series

However the researchers emphasized that ChatGPT was also able to provide at most one crucial idea which was “new and clinically relevant” in 88.9 per cent of their answers.

The results also beat that of PubMedGPT which is a unique artificial intelligence robot that was specially trained in biomedical research, that scored 50.8 percent of the test of questions based on the USMLE.

The researchers behind the study said that the research results indicate that ChatGPT could be a useful tool for medical education. “The AI bot has a limited ability to instruct in medical science by revealing new and undefined concepts that may not be within students’ heads. AI technology has poised to be commonplace in clinical practice in the near future with a myriad of applications in all health care industries.”

İlayda Mete

I am a 23 year old technology lover. I share with you the developments about technology that excite me and my experiences with games.

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button