<i>Editorial Commentary:</i> Chat Generative Pre‐Trained Transformer (ChatGPT) Provides Misinformed Responses to Medical Questions

Abstract

Surgeons have dealt with the negative effects of misinformation from "Dr. Google" since patients started using search engines to seek out medical information. With the advent of natural language processing software such as Chat Generative Pre-Trained Transformer (ChatGPT), patients may have a seemingly real conversation with artificial intelligence software. However, ChatGPT provides misinformation in response to medical questions and responds at the reading level of a college freshman, whereas the U.S. National Institute of Health recommends medical information be written at a 6th-grade level. The flaw of ChatGPT is that it recycles information from the Internet. It is "artificially intelligent" because of its ability to mimic natural language, not because of its ability to understand and synthesize content. It fails to understand nuance or critically analyze new inputs. Ultimately, these skills require human intelligence, whereas ChatGPT provides responses that are exactly what you might expect-artificial.

Ähnliche Arbeiten