ChatGBT passed medical specialty exam: will there be doctors with artificial intelligence?

ChatGPT has passed the grueling exam required to practice medicine in the US. Researchers rated this success of the artificial intelligence chatbot as a milestone in medical science.

The chatbot ChatGBT passed between 52.4 percent and 75 percent across different sections in the three-part Medical Licensing Examination (USMLE) in the US. To pass the exam, you need to get an average of 60 percent.
Researchers from the technology company AnsibleHealth, which conducted the study, said in a statement on the subject: “Achieving the passing score on this notoriously difficult expert exam and doing it without any human supplementation marks a remarkable milestone in the development of clinical artificial intelligence. used the phrases.
The results of the study have been published in PLOS Digital Health, a peer-reviewed scientific journal. However, ChatGBT had previously been tested and passed in business (University of Pennsylvania’s Wharton School of Business) and law (University of Minnesota) departments.
In the latest study, the researchers tested the software on 350 questions from the June 2022 USMLE. Two doctors evaluated the results and any discrepancies were examined by a third expert. The exam evaluates the knowledge of medical students and junior doctors in the medical field and has been in use since 1992.
Medical students in the US usually take step one of the USMLE at the end of their sophomore year of medical school, step two in their fourth year, and step three in their first year of residency after they graduate from medical school. More than 100,000 students and PhD candidates take the exam every year.
On the other hand, the researchers emphasized that ChatGPT also generated at least one key insight that was “new and clinically valid” for 88.9 percent of their responses.
The results also surpassed the performance of PubMedGPT, a distinctive artificial intelligence robot specially trained on biomedical field literature, which passed 50.8 percent on a USMLE-style exam.
The study authors stated that the study results show that ChatGPT can become a valuable tool in medical education: “The AI ​​bot has a partial ability to teach medicine by exposing new and unclear concepts that may not be known to the students. are. Artificial intelligence technology is now positioned to soon be ubiquitous in clinical practice, with a variety of applications across all healthcare sectors.”
ChatGPT, recently acquired by Microsoft, is a large language model that uses a variant of the Generative Pre-training Transformer (GPT) architecture to generate human-like text. The chatbot is trained on a variety of internet texts and is capable of generating substantial responses to a wide variety of commands. The model can be refined for different tasks, such as translating languages, answering questions, and speaking. ChatGBT was created by OpenAI’s research team in the USA and is constantly updated based on new research and developments.
Q&A Writing text (basic academic articles, literary texts, film scripts, etc.) Solving mathematical equations Debugging and correcting (e.g. finding and correcting errors in a block of code) Translation between languages ​​Summarizing text and identifying keywords in the text Making recommendations Classification Explaining what something does (e.g. what a block of code does)

Leave a Reply

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