^ Gee, for real? 織日ちひろ?
Okay. The way you use it is much better than just treating LLM's output as truth, which was what I intially imagined. I'm sorry for overreacting whenever I saw people using LLMs when I didn't know how they used it.
Though, still, since you have doctor friends to ask about these questions, maybe you can just ask them for titles of reference materials you can check? (maybe you can even borrow their books) I'd think doctor friends are much more reliable in medical information than stocastic parrots, as they often give out unreliable sources or made up nonexisting books. Also, aside from hallucinations (making things up) and sycophancy (affirming whatever users give), there are also grave ethical concerns regarding LLMs' training and inferencing process, like energy cost, water usage, web scraping that's basically DDoS attacks, and the pollutions generated by AI companies building up fossil fuel power stations. Microsoft even plans to restart Three Mile Island nuclear power plant, the scene of the worst commercial nuclear accident in US history, just to power its AI.
water and energy usage: 1 2
web scraping: 1
Three Miles Island: 1
And, I like your work very much. Do you have something like patreon or ko-fi?