As per Moneycontrol, ChatGPT, a big language model built by OpenAI, barely passed all three sections of the United States Medical Licensing Examination (USMLE) in a research trial.
The United States Medical Licensing Examination (USMLE) is a standardized exam that evaluates the knowledge and skills of medical students and graduates and is required for licensing to practice medicine in the United States.
The examination consists of three distinct sections: Part 1, Part 2 Clinical Knowledge, and Part 2 Clinical Skills. Typically, second-year medical students spend hundreds of hours preparing for Part 1, while medical school graduates take Part 3.
Ansible Health, a Silicon Valley firm specializing in the treatment of Chronic Obstructive Pulmonary Disease (COPD), had been investigating various AI and machine learning techniques to enhance its care before deciding to test ChatGPT, the latest AI craze.
According to them, ChatGPT was trained on a massive amount of medical material and exhibited a profound understanding of the topic. It was capable of providing precise responses to a vast array of medical concerns, including those pertaining to the diagnosis, therapy, and medical ethics.
Prior to administering the test, the staff confirmed that “none of the answers, explanations, or related content were indexed by Google.” This has major ramifications for the future of medical education and healthcare as people continue to consult Google for health-related inquiries.
ChatGPT and other similar models could potentially be used to assist medical students in their preparation for the USMLE and other medical examinations, as well as to aid physicians and other healthcare workers in their daily job.
However, the research is still in its infancy, and additional study is required to determine the full potential of the technology. However, ChatGPT is not (yet) capable of replacing human physicians, and traditional medical education and practice will continue to play a crucial role in the diagnosis and treatment of patients.