Korean J Med.  2023 Jun;98(3):99-101. 10.3904/kjm.2023.98.3.99.

In the Era of ChatGPT, Can Medical Artificial Intelligence Replace the Doctor?

• Kim S

• KMID: 2542685
• DOI: http://doi.org/10.3904/kjm.2023.98.3.99

Abstract

Whether artificial intelligence (AI) can replace the role of doctors has seriously been discussed since the appearance of Chat Generative Pretrained Transformer (ChatGPT). Unlike past expectations that limit the role of AI in the medical field, doctors seem to both admire and be threatened by recent advances in AI models. Because previously developed and approved AI models have been managed and reviewed by medical experts during the preparation, refining, annotation, and verification of data, those systems can be considered to be based on evidence-based medicine. However, the current version of the ChatGPT model derives the most meaningful results from unverified open data. This approach enhances the accessibility to new information but is significantly different from the methodology of evidence-based medicine. Like all AI models developed to date, ChatGPT needs a system that can be rigorously verified and regulated by doctors to facilitate its use in the medical field.

Keyword

Artificial intelligence; ChatGPT; Evidence-based medicine; 의료 인공지능; ChatGPT; 근거중심의학