The follow-up post asking ChatGPT which model it's based on is... dubious. Yes, OpenAI may have added that information in the system prompt because they know people keep asking these things but people forget the LLM will answer based on the <i>training dataset</i>. You can't rely on super accurate information that way, even if it's browsing the web for an answer (it may find the wrong article and give you outdated information).