Logos AI is very limited. Are there any trusted ones? Any we can integrate?
you can not trust an AI - in general. they can not think, they can not comprehend, they can not feel, they can not understand. What they do is: based on (human) training data: give you the statistical most probable combination of words. And they do it quite good - so good that we often forget that this is not actual intelligence. It is an software program that imitate human intelligence. So how could you trust it?
I have been using ChatGPT and Grok. (Free versions of each). I got the impression that ChatGPT is better overall, but Grok was better at things requiring math and logic skills.
I started using Googles AI a little bit. Not sure if what you get in Google search is the same as Gemini. But it seems to be on a higher level. Much more like having a conversation with a real person. And it apologizes a lot less after you find it giving a wrong answer.
This makes me wonder how much better the more advanced models might be.
I have not really used Logos AI much. But the times I did it did not utilize my library in a way that it would give better answers than the mainstream AI, even on Biblical and Theological questions. What Logos does better is it gives references immediately. That is good. What is bad is that it does not select the best resources.
As far as trusting AI, any AI, I think it is like talking to anyone. If it isn’t God speaking, then it is a fallible source. You always verify anything that is important.
I understand that Logos’s Study Assistant searches my library for answers and commentary based on the prompts I enter. It usually finds relevant material and cites it correctly, though the responses often feel surface‑level and could draw more thoroughly from my resources. I’ve also noticed that it sometimes misinterprets what the cited book actually says. When I point this out, it typically replies, “You’re right; I misinterpreted the author,” and then launches into a “cleanup on aisle 9” routine. Because of this, I remind people that the “Early Access” label is a caution flag. “Trust, but verify” has always been my motto for AI.
I'll be frank and say that I cringe every time I see you write this about AI ("Trust, but verify") because my mind always goes, "No, no, no, never trust, and always verify." Sorry, I had to get this off my chest 🙈
That's the point.
The problem is, since the AI mimics a human quite good, the people trust the AI far more than they should.
I wrote myself some regulations for the AI use in context on God's word, printed it, and hang it above my PC on the wall. So whenever I study I remind myself, that IF I use AI, I'll use it only for research on non spiritual topics.…
Ha! I totally understand. The saying became widely known because Ronald Reagan used it frequently in the 1980s during nuclear arms negotiations with the Soviets. It stuck, and it’s been a common everyday saying ever since. I agree with you, though, regarding the trust aspect when it relates to AI.
What I like about using the AI functionality is Logos is not the answers it gives, but the sources it points me to. I'm especially looking forward to the ability to use collections in the Study Assistant, so I can move beyond the five sources limitation and make fuller use of my library. I see lots of new collections in my future 🙃
I use 6. You get the hang of which is best at this or that. Sometimes I just enter the same prompt in all six and compare results. I could do that with a Python script, but I am too lazy. Yes, my trust level is very low. But like the news—-trust level is close to zero, but we glance at headlines anyway.
i appreciate that approach of logos, i also like that. i absolutely don't want to know what any AI says to a spiritual topic - so i like that it is giving me book quotes instead. what i don't like, is the AI in the sermon assistant and on other places. When you let your AI doing the outlines - than you shouldn't do a sermon, because you did not understand the text. Same for questions, if you are to busy/lazy to think about own questions, you should not have any ministry. I fear that using AI makes us more lazy, less disciplined and even more hypocrite. More knowledge alone will not make you a better diciple…
AI can improve your outline after you understand and create it. It can help build better thinking and better habits. It can check for errors—many kinds, including theological. It is great for finding quotes (double-check all sources). Just double-double check anything it tells you.
No.
James, we'd love to learn more about what limitations you have in mind—what you wish Logos could do but currently doesn't.
We're very intentionally not trying to create a direct competitor to or replacement for ChatGPT, Gemini, Claude, Grok, et al. Rather than focusing on machine-generated content, we're leveraging the advances that GPT LLMs enable to help you find and engage with the human-authored content in the Logos library.
Available Now
Build your biblical library with a new trusted commentary or resource every month. Yours to keep forever.
