JMIR Medical Education | How does ChatGPT perform on the USMLE?
In this video, Gunther Eysenbach, chief executive officer of JMIR Publications and executive editor of the Journal of Medical Internet Research, interviews Dr Andrew Taylor from Yale University School of Medicine on their recent publication highlighting the potential of ChatGPT in medical education. Dr Talyor and colleagues examined how the AI model ChatGPT fared on the United States Medical Licensing Examination (USMLE)—read their full paper published in JMIR Medical Education (https://doi.org/10.2196/45312).
In their study, Dr Taylor and colleagues compared ChatGPT to other large language models, such as InstructGPT and GPT-3, and found that ChatGPT's additional dialogic component, driven by prior training on question answering data sets, resulted in additional information and reasoning for the answers it generated. They used several data sets reflective of questions on the USMLE and found that ChatGPT's accuracy in answering those questions was at par with a third-year medical student.
In the interview, Dr Taylor discusses the potential for ChatGPT to be used as an AI-based tutor system or peer buddy for educational purposes, as well as ChatGPT’s limitations, such as the need for structured prompts. Read Dr Gunther Eysenbach’s editorial for an interesting conversation with ChatGPT on this topic.
Watch this video interview to know more about ChatGPT's potential use in medical education and how it could revolutionize the way we learn and prepare for exams.
Submit your research on the role of generative language models and AI in medical education to JMIR Publications: https://mededu.jmir.org/announcements...
###
DOI - https://doi.org/10.2196/45312
Full-text - https://mededu.jmir.org/2023/1/e45312
Corresponding author - David Chartash, PhD, Section for Biomedical Informatics and Data Science, Yale University School of Medicine, 300 George Street, Suite 50, New Haven, CT, US
Phone - 1 203-737-5379
Email - [email protected]
Keywords - natural language processing, NLP, MedQA, generative pre-trained transformer, GPT, medical education, chatbot, artificial intelligence, education technology, ChatGPT, conversational agent, machine learning
About JMIR Publications
JMIR Publications is a leading, born-digital, open access publisher of 30+ academic journals and other innovative scientific communication products that focus on the intersection of health and technology. Its flagship journal, the Journal of Medical Internet Research, is the leading digital health journal globally in content breadth and visibility, and it is the largest journal in the medical informatics field.
To learn more about JMIR Publications, please visit https://www.JMIRPublications.com or connect with us via:
YouTube - / jmirpublications
Facebook - / jmedinternetres
Twitter - / jmirpub
LinkedIn - / jmir-publications
Instagram - / jmirpub
Head Office - 130 Queens Quay East, Unit 1100 Toronto, ON, M5A 0P6 Canada
Media Contact - [email protected]
The content of this communication is licensed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/..., which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, published by JMIR Publications, is properly cited.
