This past Sunday we were talking about things in the Bible that were literal vs. things that were idioms, and I mentioned that we say Jesus was resurrected "three days later" because that's how the Jews said it. They referred to any part of a day as a full day's time, so Friday night to Sunday morning was called "three day" instead of two days or 36 hours. Someone in my class said that Jesus was buried for three full days because of the passage saying he would be there the same as Jonah was in the belly of the fish, so the Friday death is tradition and he actually died on Thursday. He then sent me this conversation on ChatGPT where he digs into it. https://chatgpt.com/share/69788824-6c64-800e-a65b-66161b130520 I had done a search in Logos using the new AI feature, and I found the opposite. My resources said he was definitely buried on Friday. I think this ChatGPT conversation is very well done, so I want to put my friend's prompts into Logos, and I want to try my Logos prompts into ChatGPT. To be clear - I'm not trying to prove my friend wrong or show that I am right - rather, it's very curious to me that his AI engine told him that he was correct, and my engine told me that I was correct. Is that AI bias trying to please the user? Or did we each phrase our questions in such a way that helped the answer come back in our favor? So my experiment is just as much about learning how the AI engines work as much as it is finding the correct answer. (My friend is much more knowledgeable about Scripture than I am, in my opinion, so it's certainly possible that it will be me that learns something new here). So here is my question - how do I find my past AI search in Logos? In Tools / Library / History I can see words I've searched for, I can see where I've been to help articles and Study Assistant this morning looking for this answer, and I can see where I researched the word 'Mercy' to prepare for Sunday's lesson. I don't see any AI searches at all. I am newly on Premium, if that matters. Does anyone have a suggestion?
ChatGPT - Sign of Jonah vs Tradition
Shared via ChatGPT
chatgpt.com

  2. When I use ChatGPT, I set the parameters as an expert in Reformed theology and then I specify what Bible translation to use. I found that the responses coincide with the AI in Logos. If you don't set any parameters, then the ChatGPT will respond in a more liberal fashion.

  3. There is a post on X showing that chatgpt, gemini, and grok all show bathing suit photos of women on command. From my perspective AI is just that: artificial. I've written a lot of software over a long time and it's true that GIGO (garbage in garbage out). Stating that "At least ChatGPT doesn't undress photos of young girls" is disingenuous at best. The big problem everyone is letting themselves in for is that AI, though artificially intelligent, has no mind. Christians have the mind of Christ and should use AI with a great deal of caution, always checking results for falling off the rails because guardrails are almost non-existent.

  4. Actually, ChatGPT has very tight, multi-layered safeguards around sexual content, especially anything involving minors. You generally hit hard refusals early, even with indirect prompts or image manipulation attempts. Grok, especially in earlier releases, was designed to be more permissive and less interventionist. That reduced friction made it easier for users to push into disallowed territory before safeguards kicked in.