loading page

Artificial Intelligence Chatbots in Medical Education, Opportunities and Challenges: A Quantitative Analysis
  • Mikaela Koch,
  • Nishanth Vemuri,
  • Aparna Sridhar
Mikaela Koch
University of California Los Angeles
Author Profile
Nishanth Vemuri
University of California Los Angeles
Author Profile
Aparna Sridhar
University of California Los Angeles

Corresponding Author:[email protected]

Author Profile

Abstract

ChatGPT is an artificial intelligence chatbot released in 2022 marketed as a powerful tool for education. The latest version, ChatGPT 4, has performed well on standardized exams, passing all three United States Medical Licensing Examination (USMLE) Step exams 1. Compared to its predecessor (Version 3.5), the subscription-walled Version 4 has been trained with roughly 500 times more parameters and has been advertised as a much more efficient model. Given the increasing utility of chatbots in our society, further investigation is required to confidently assess ChatGPT’s efficiency as a learning resource in medical education. In this study, we explored opportunities and challenges presented by chatbots by assessing ChatGPTs accuracy in answering National Board of Medical Examiner (NBME) style questions in OBGYN.