HCP Live April 24, 2024
Article intelligence chatbots, including ChatGPT and Bard, varied in accuracy, comprehensiveness, and readability on glaucoma questions adapted from patient brochures.
A new evaluation measured the strengths and limitations of the responses of multiple artificial intelligence (AI) chatbots, including ChatGPT, Bing, and Bard, in answering glaucoma-related patient questions.1
Overall, AI chatbot response accuracy was below that of the American Academy of Ophthalmology (AAO) glaucoma-related patient education brochures. However, the accuracy scores of Bing and ChatGPT were only slightly behind the handout, suggesting their utility for glaucoma patient education.
“With an estimated 43% of glaucoma patients utilizing the Internet for medical information, it is essential for physicians to understand both the information AI chatbots provide and how it is provided, such that...