Exploring Artificial Intelligence Programs' Understanding of Lens, Cataract, and Refractive Surgery Information.

Abstract

Purpose: We aimed to evaluate the success of Chat Generative Pre-trained Transformer (ChatGPT), Bing, and Bard artificial intelligence programs, which were released free of charge by three different manufacturers, in correctly answering questions about lenses, cataract, and refractive surgery, and to investigate whether they are superior to each other.

Methods: Seventy questions related to lens, cataract, and refractive surgery were obtained from the study questions section of the American Academy of Ophthalmology 2022 to 2023 Basic and Clinical Science Course Lens and Cataract and Refractive Surgery Books. The questions were asked separately for the ChatGPT, Bing, and Bard artificial intelligence programs. The answers were compared with answer keys and grouped as correct or incorrect. The accuracy rates of artificial intelligence programs were compared statistically.

Results: ChatGPT, Bing, and Bard chatbots gave correct answers to 35 (50%), 43 (61.4%), and 50 (71.4%) questions asked, respectively. The rate of correct answers to the questions of the Bard artificial intelligence program was significantly higher than that of ChatGPT (P = 0.009, Pearson's Chi-square test).

Conclusion: Although the currently released artificial intelligence chatbots can be used to access accurate information about lenses, cataracts, and refractive surgery, one should always be careful about the accuracy of the answers given.