We found the article “Performance of Generative Large Language Models on Ophthalmology Board Style Questions1 interesting. Cai et al look into how well generative AI models can respond to questions similar to those on an ophthalmology board.1 When human respondents are used to answer questions from the BCSC SAP, LLMs , in particular ChatGPT-4.0 and Bing Chat, can perform comparably, according to Cai et al.1 The incidence of hallucinations and non-logical thinking suggests that conversational agents’ performance in the medical sector should be improved.