ChatGPT, the Conversational AI Robot, Approaches Medical Exam Pass Rates

Have you ever imagined a robot that can take a medical exam and score better than most students? Well, that is precisely what ChatGPT, the conversational AI robot, has accomplished, according to a study published in PLOS Digital Health.

The study was conducted by AnsibleHealth, and it tested the software's performance using the United States Medical Licensing Examination (USMLE). This exam evaluates students in various domains, from scientific knowledge to clinical reasoning, and it is divided into three parts.

The researchers tested ChatGPT on 350 of the 376 questions published on the USMLE website and included in the June 2022 exam. The questions were presented in three formats: open-ended questions, multiple-choice questions without justification, and multiple-choice questions with justification. However, the questions based on images had to be removed.

The software received scores ranging from 52.4% to 75%, approaching the passing grade of 60%. The results were encouraging, suggesting that AI has the potential to assist doctors in diagnosing and treating patients.

Despite criticisms of the study's methodology from some external experts, who suggested anonymizing responses from humans and robots, Nello Cristianini, a professor of AI at the University of Bath in the UK, praised the research as an exciting development in the field of AI.

Lucia Ortiz de Zarate, a researcher at the Autonomous University of Madrid, believes that this study demonstrates the potential of AI in the medical field. "The software can be of great help to doctors when it comes to formulating diagnoses and prescribing treatments," she said.

While the future of AI in medicine is uncertain, the results of this study are undeniably impressive, and it is exciting to see how technology can help improve the healthcare industry.

Source : ici Radio Canada