Sometimes a question leads somewhere unexpected. This article started as a simple look at a Barna study, but the conversation that unfolded became something I believe is more valuable: a real‑time example of how a Christian can use AI with discernment, curiosity, and spiritual grounding. I’ve left the dialogue mostly as it happened – pauses, surprises, “tech‑shed” moments and all – because the process itself is the lesson. My hope is that this gives you a safe, honest look at how AI can be a tool without becoming a spiritual authority.

To keep it “real”, what follows is the initial introduction, and then you can see exactly how it all changed through the “dialog” between me and Google Gemini for the most part – with a small exchange at the end between me and MS Copilot.

BTW – the adjacent image may or may not be possible today, but (sadly) it’s not my home setup.

---

Barna research recently published “AI is Becoming a Spiritual Authority, Even Among Practicing Christians“. That’s potentially a scary thought. As usual, Barna published a lot of numbers. There’s some analysis, but not that much. And they rarely reach conclusions as to the pros and cons. But I will.

Although I use AI in some of my research, even I wouldn’t say I trust it. Not for spiritual authority or any other authority. That’s just plain dangerous. It almost reminds me of, “Did God really say …?”

Not that AI is from the devil or anything like that. AI can be very useful. But trustful? How can you really trust something that keeps repeating “AI can make mistakes” frequently while it’s responding to our queries?

The Barna article is rather telling about the conflicting attitudes behind their title. You see, the title seems to say that even practicing Christians consider AI to be a spiritual authority. And yet, the article also says:

✏️ Practicing Christians are willing to trust AI with their spiritual growth. They are also deeply worried about what happens when AI starts acting like a spiritual authority. New research from Barna finds that the Christians most open to AI are often the same ones most unsettled by it.

Barna Research

That quote indicates a conflict that’s not obvious from the title.

What is this seeming conflict of AI trust about?

Going a bit further into the article, we see more of what’s happening.

✏️ The worries are not abstract: They cluster around scripture, divine authority, and the integrity of faith itself. Nearly three-quarters of U.S. adults (74%) express concern about AI misinterpreting scripture. Among practicing Christians, that figure rises to 83 percent, and among pastors, to 94 percent. Nearly two in three practicing Christians (65%) express concern about AI beginning to act as a replacement for God, compared to roughly four in five pastors (79%). Seventy-two percent of practicing Christians are concerned about AI replacing the role of pastors or spiritual leaders, as are 63 percent of pastors. And nearly three-quarters of practicing Christians (73%) express concern about people losing their religious faith because of AI.

Barna Research

Those numbers don’t really look like there’s a lot of trust for AI when it comes to spiritual authority. There’s a high level of distrust.

The issue comes, in large part, because of the question driving the headline. That question was: “How much would you, personally, trust AI’s advice in these areas? One of those areas was “Growing Spiritually”? The numbers came out:

Group	% Completely Trust + % Somewhat Trust
Pastors	12%
Practicing Christians	48%
Non-Practicing Christians	38%

You can see two issues here

The obvious one is that they combined Completely Trust with Somewhat Trust and wrote about it as “willing to trust”. I must say, willing to trust and completely trust are two very different things.

The not so obvious one is how Barna differentiates practicing and non-practicing Christians. It’s not in this article specifically (that I noticed), but Barna often states their underlying classification of practicing Christians. In some cases, someone who attends a church service on Christmas and Easter only – twice a year – is counted as a practicing Christian.

Those two things can make huge changes in the final numbers and therefore the conclusions reached. I’ll be writing more on this later in additional articles that examine the details.

Should Christians “trust” AI?

As you may have noticed, I use AI in both of my sites. I have special formatting for AI responses that are included in my articles. They identify it as AI produced and which AI generated that content – usually Faithlife/LOGOS for research via their extensive eLibrary of various Christian books, Claude, MS Copilot, and Google Gemini. I also verify that content rather than blindly accept it, push back with that AI if there are problems, and add my own comments to it.

I also have categories on both sites that are dedicated to “Conversations” with AI where I include an entire session, going back and forth with the AI agent, that ultimately is a large part of my initial research for some of the more complex articles. I do this with the hope that it will help show the benefits and the things to watch out for when using AI, especially for spiritual and theology questions.

If you’d like to check them out, here are the links to those articles on both sites.

Conversations With AI – Which God Can Save? (this site)

Conversations With AI 🤖 – God Versus Religion (deeper into Bible Study and theology)

To that end, after reading this Barna Research article, I decided to have a “conversation” with Google Gemini about this issue – and this article.

An AI Conversation with Google Gemini about Spiritual Trust and AI

Here’s the article, in its entirety, along with some comments from me about what you’re reading – what’s good and what we need to be careful with.

---

Should I trust AI, according to Google Gemini?

---

There are some things to note here.

1. within my specific limits as an artificial intelligence: This is true. However, few people understand the limits of artificial intelligence. Plus, they change with each release. Gemini goes on to break out 4 areas of trust, but even there – the answers are within its limits, as we’ll see.
2. Humanity (Intent & Empathy): No, do not trust me here: Absolutely true. Watch out here though.
   • I’ve actually had MS Copilot respond to me once with something that started with it telling me it was going to explain something “human to human”. Say what?
   • I’ve also had AI responses telling me they “understand my frustration”. They do not understand!
3. Capability (Competence): Yes, with verification True – and listen to this one: always verify.
4. Transparency (Honesty): Yes. Sort of.
   • The level of honesty is really controlled by its “training” – what’s fed into it as sources for possible responses.
   • It’s also controlled by the programming – up next,
   • If you pay for the AI, or if you enable “memory” on even a free one, it begins to focus on responses that align with previous questions from you and its responses to you. That means, over time, your responses will have an increasingly narrower focus. If you want something “different” then you’ll need to specifically ask for it. Examples of this will come up shortly.
5. Reliability (Consistency): Yes, within my programming Sorry, but not entirely true.
   • Yes, there is the programming part. Think of that as being like the ads you see on social media and even many web pages. They track the things you click on, the things you disable, and try to focus in on more things you like.
   • AI can have “bad days” – after a fashion. AI used to “hallucinate” a lot – some up with responses that had nothing to do with the question, or sometimes that are downright dangerous. It happens. With some of them, the longer any one “conversation” is, the more likely it’ll go off the deep end. It especially happens when the topic is too wide or when conflicting thoughts are brought in and not clearly delineated. And remember – that can be the way you present it or the way its programming “read” it. When it goes off like that, just start a new session.

I almost want to say trust but verify – but that’s too simple. Sometimes we know right away that something’s off. But other times, even with verification, if we look up something we don’t know – both AI and us can hit sites that are wrong, off, misleading, and we might not catch it.

So just know that some level of verification is always required – whether it’s confirming something we already know – or spending lots more time doing research. There’s also the possibility of asking 2 or 3 AI’s the exact same question.

---

Does Gemini think I trust it for spiritual authority?

Let’s move into spiritual authority questions next. Since I use Gemini for some of my research, we have a “history” that it “remembers” – as mentioned above. So, I asked a question based on bringing in that history. Based on my question, it knows I’m focusing in on its question back to me: “What is the specific angle of your piece?”

---

Clearly, Gemini doesn’t “understand” my level of trust – or should I say, mistrust?

1. you trust my capacity to maintain context and process complex ideas: As you just read above – no, I don’t trust this part at all. I’m always watching for any AI to either (1) go off the rails completely or (2) get sidetracked by a typo on my part, a misunderstanding anywhere, or who knows what. Yes, it almost always returns something that sounds authoritative. Even after being corrected several times, it always comes back with something along the lines of, “This is the real answer” – when in fact the longer that goes on the less likely it’s going to get the real answer.
   Once again, when you reach this point, start a new session from scratch, because that one’s not going to get better.
2. trust my programming to handle your ideas without deleting or distorting them: I worked in IT for more than 35 years, so I know better than to trust any programming. I also know from experience, asking AI to write code for me, it’s not unusual to have to go back and get it to fix problems. And, I have to add, it’s quite annoying when I’m using its code and it tells me what’s wrong with My Code! At least it could take ownership of its own issues, but it doesn’t. 🙄
3. you trust that my architectural training will yield useful, coherent linguistic patterns in return: As mentioned earlier – it always sounds authoritative, even when it’s incoherent. And even when it’s wrong. I guess that means I don’t really trust this either.
4. In the context of spiritual authority, your trust in me naturally hits a hard ceiling: Absolutely. And the ceiling is touching the floor. We’ll get more into the “why” of that in a bit.
5. You trust me to act as a mirror and a filing cabinet for your thoughts, but you do not trust me as a guide: Mostly.
   • AI acting as a mirror, as implied earlier, can be good and bad.
     • If you use AI and don’t allow it to remember, or don’t sign it, then it’s not a mirror or a filing cabinet. Each interaction is like two people who just met for the first time.
     • If you do allow AI to “remember”, there are the caveats above related to what that does – narrowing its focus to try to match yours. The next thing you know – it can almost turn into a parrot. That’s not good. The only way to get a varied point of view on some things is to ask for it. Not that this is necessarily a bad thing. It’s just that we need to be aware of it – ask accordingly, push back, give it a different direction, etc.

I know – this may be getting a bit more complex than you expected. Sorry, but it’s not going to get any easier. The question is whether or not it’s worth the effort to use AI properly.

You’ve probably read, seen, or heard various ways to “format the best prompt”. When you do things like this, forget all that. Knowing full well AI isn’t human, remember – it is programmed to act like a human in many ways. It does a lot of pattern matching. So, when you ask a question, ask it the same way you’d ask another person. You’ll see that in this article, and especially in the AI conversations links if you go to either of those.

Try to keep that in mind as we move along. It should make the task of using AI to get useful information more simple.

---

Does AI challenge traditional definitions of discernment?

Next, let’s take up Gemini with a variation of its suggested follow-up topic: “Would you like to explore how AI challenges traditional definitions of “discernment” in spiritual reading?”

I have to give a warning about this answer. I’m going to really push back hard on this one. I’m leaving it in, because it is part of the conversation. It also shows the need to absolutely step in if something’s off. The way I push back is also important – because I want to be as sure as possible that it gets the message on why I objected – and to direct it to regenerate a new reply with the correction included in it.

---

This is just so wrong! And it all comes down to one critical error. Sadly, it’s an error that many people will probably think is awesome. But – it’s not.

1. we must shift from evaluating the source of a message to evaluating the fruit of the message and In short, AI strips away the luxury of relying on external spiritual authorities. By presenting flawless spiritual language from a soulless source, AI forces the human reader to take 100% of the responsibility for discernment back into their own conscience: Not at all. It’s exactly the opposite.
   • The source is always important. There’s a reason why there are so many denominations of Christianity. Each group that splits off has reasons for why they broke off from the old denomination. While some may appear relatively minor to the uninformed observer, something had to be pretty major in the eyes of the group that broke off.
   • Some of the denominational differences are major. You’ll see in a little bit how AI deals with that, but trust me (no pun intended), it’s not a solution that should make anyone happy.
   • It’s interesting that it chose to remove the source as important – and that the fruit should be examined. There are plenty of examples in the Bible about false Apostles and teachers, claiming to be doing God’s work but in reality God has nothing to do with it, and the like. Jesus told one group – I never knew you. In Revelation, Ephesus was commended for testing and finding false Apostles, but right after that was told they were about to lose their status as a church because, even though they were doing a lot of the “right” things, they were being done without Love – in other words, without God.

We can’t remove the source from the message without losing the weight/trustworthiness of the message!

---

And so, here’s the pushback, and a further look at what went wrong with the first answer.

So, we have in that question:

• The push back
• The reason – sources are important
• A reason why the sources can be even more critical when interacting with AI
• A real-life example of a problem that cannot be resolved without the sources
• Closing with the reality that it was only one example of many possible scenarios where sources are important.\

---

For this one, think back as far as you can (varies depending on your age) and remember how you used to do research. I remember going to the library and using the Reader’s Guide to Periodical Literature to find relevant magazines & newspapers, as well the Card Catalog to find books, on whatever topic I was researching. There wasn’t a computer to be found – anywhere – at that time.

Then, when I started working, we had punch cards (Google it) and the beginnings of computers that took up an entire house-sized room, but would be put to shame today by the average watch.

My point is, besides that in spite of this, my mind still works – is that I remember the days when sources were completely transparent and obvious. And relied on. Now, sources can be largely hidden – a sad fact when we realize the very same technologies put out more and more information that’s utterly useless – and incorrect.

With that in mind, let’s look at the bullet points for this critical feedback section.

1. I missed a critical layer of how AI actually operates, and your example about Faithlife and Calvinism perfectly illustrates why: As I wrote, as did Gemini – AI can make mistakes. This is proof that, unsurprisingly, it does indeed make mistakes.
   • My previous point about shifting away from “Who said it” only applies to the front-end output (the text you read), but it completely ignores the back-end lineage (the training data): – this is a real issue. I know this, because I worked in IT for a long time, and use it extensively in my writing. In other words, I have a background to understand the AI side of things that many people don’t have – probably especially for people who write about Christianity. I mean – how many religious people think AI is something to be afraid of? Even the Pope warns people about it.
     The thing is, AI – in and of itself – is like so many other “things”. AI is neither bad nor good. It is what it is. But – the choice of which AI product to use affects its underlying programming, its sources of information, its guardrails to hopefully prevent it from going off the rails, and so much more – is very much based on which “flavor” of AI you use.
     As you’re seeing, having some understanding of how to interact with it also matters.
     As does some understanding of how it works internally.
     And maybe the more dangerous part – especially evident in this example – is having enough understanding of the topic you’re researching to be able to discern when the responses are “off”.
2. AI doesn’t actually eliminate the need to look at sources—it drastically raises the stakes for doing so. Here is how your insight reshapes the concept of discernment: This flips the AI presentation method on its head. That’s a major piece missing. Let’s take another example – If you have a really strong pain in your side – would you want an AI answer to come from the website of a doctor who lost their license for incompetence or malpractice? I sincerely hope not!
   Isn’t your eternal soul – or the souls of your friends and loved ones – worth taking the time to be sure you’re getting correct/calid answers to your questions?
3. The Invisible Lineage: This goes to my comments about Reader’s Guide to Periodical Literature and Card catalog drawers. You look up physical books, magazines, newspapers, etc. They say right on them who the author was. Often, there’s a short bio so you can learn more about who wrote what you’re reading. Learn their biases. Their doctrine they follow. All sorts of clues as to the validity of what you’re going to read. And consider this – with religions, theology, doctrine – each one believes they got it right. But, they don’t agree. So, the “right” answer may very well depend – to you – on the religion/denomination of the author. That’s easy in a paper source. It can be very difficult to impossible with AI.
4. The Marketplace and the Algorithm:
   • Part of this is the ever-narrowing search window AI will search as you use it more often. Knowing this – and asking questions in a way to keep this window open is critical. Earlier, Gemini referred to itself as a mirror. Eventually, you may need to ask yourself – do I really want to have AI feed me what I already believe, when I’m trying to learn more about something?
   • It also goes to knowing the “training” of the model. AI cannot tell you anything it hasn’t been trained on. So, if you’re trying to find out something about a niche category, it just plain may not know. Then you get nothing more than a web search – and who knows what that might give you? That scenario is probably why you went to something that’s “intelligent” – even if it’s artificially intelligent. BTW – that means it appears to be intelligent, but it’s not intelligent in the way we are.
5. This means that using AI for spiritual research actually requires more traditional theological expertise, not less: This is probably one of the more discomforting statements. It’s not easy to use AI for theology. Especially for a beginner. You need to have a way to determine the validity of everything AI responds with – a friend, pastor, study Bible you trust, web site you trust, someplace that you can go to for some basic level of “truth” about what you read as it relates to what the Bible says and means.
6. Instead of AI making the “source” irrelevant, AI anonymizes the source, making the human’s job of unmasking and weighing that source far more difficult—and far more critical—than ever before: This is where it really gets messy.
   • Earlier, I mentioned we’d get into what can happen when AI puts together answers with no sources. This is the time. Let’s say you’re Catholic, but you don’t mention that. AI looks up your question. It mashes together information from various Christian sources – which may or may not even include a Catholic source – and gives you an answer. But – does it line up with your Catholic doctrine? Maybe – but maybe not.
   • As an example, I asked a question a while back and got an answer from the book of Jubilees. I’ve never heard of it. Turns out, it’s most often used by an Ethiopian denomination of Christianity. That honestly doesn’t help me.
   • Or – another one referred me to the Book of Enoch. Sorry – that’s not a book in the Bible my denomination uses either.
   • Sometimes these things matter. Sometimes they don’t. But ir we don’t know where the answer comes from – we also don’t know whether or not we should trust it in our context, even though it’s valid in a different context.

To summarize – very briefly – Be Careful!!

Time to completely change the direction of this “conversation”.

---

My use of AI for this article on the Barna Research article

Now, I wanted to turn the answers back to the Barna article that got me started with this line of questioning in the first place.

---

This is mostly about how I use AI, so the explanation is like a summary of what I often do.

1. By analyzing the data you referenced alongside your own multi-model collaborative research method, we can map out a robust chapter or article section on “The Discernment of Data.” By multi-model research, Gemini’s referring to my overall approach.
   • My own background, after 15+ years of research, teaching Bible Study classes, and writing for the web.
   • I have an eLibrary of 5,000+ books from Faithlife/LOGOS, all fully searchable by content, person, event, date, etc.
   • Faithlife/LOGOS AI agent with access to the entire LOGOS library, but only summaries for books I don’t have a license for. If nothing else though, the summaries are good for further research ideas.
   • In addition to Google Gemini & LOGOS AI, I also use MS Copilot and Claude for research. When questions arise from an answer by one of the AI agents, it’s easy to ask a different one the same question – then proceed based on what I find.
2. Deconstructing the Barna Trust Gap: Gemini then went and did its thing to address the Barna data question I gave it. I left in the links so you can check them out.
3. The “Safe Research” Framework: A Christian Guide to AI Co-Working: Not everything is done with AI, although the response makes it seem that way. The LOGOS software package allows for doing a lot on my own – including all of the things listed that I use it for.
   Often, whether I do it myself or use AI is a question of my time. Specific, narrow questions, it’s easier to do it myself – even getting Hebrew or Greek sentence diagrams, meanings, and especially cultural and time period related nuances.

I think the bottom line is that the more we learn, the more we want to learn – at least for me. And the more I want to learn, the more resources it takes, and the more old-fashioned computer programs and now AI can help me find things faster.

Having said that – each of us should be responsible for our own faith – and not rely on anyone, machine or human, to tell us what to believe. Yes – seeing other viewpoints is important. But more important are our hearts – and that’s what God really looks at – not whether we use AI or not – and not which AI we use.

---

Tech-Shed Guide for Modern Disciples

Tech-Shed Guide for Modern Disciples? I was in a hurry and went with it. Really though – I wasn’t sure what Gemini meant by it. Was it “shed” like in a little building to learn how to use tech? Or was it shed, as in removing, tech from our study? Talk about getting something from AI that’s not a reliable answer!

In any case let’s see what Gemini had to say.

---

You see now, we started with a general question on trust – moved the discussion to include the Barna Research that this article is about – closed the loop by getting information on how to structure an article (not this one but probably a future one) – and in the process, along the way, I made sure Gemini got straight with what I believe is the correct way to do this kind of research.

In this case, it’s not about theology – it’s about the research process.
The “Conversations with AI” categories on both of my sites though – they do get into how to straighten out the AI Agent to retrieve theology to match what I believe. Having said that – sometimes I want general research, covering multiple denominations, in which case I drive the conversation differently. It takes practice, but it can work, if we’re watchful & careful.

1. Discipleship is an information-heavy, lifelong, text-intensive process: I included this because it’s true. We’re all called to do some amount, some form of discipleship. But not everyone is called to do it this way.
   However, as someone with a lot of experience in IT and in Bible research – I do hope this helps someone who’s interested in doing the same, to whatever level you want to go.
2. 📜 The Evolution of Access: From Scrolls to Semantic Search: It’s new. AI has brought search to a whole new level. That can be good, it can be not good. Ultimately though – it’s up to us. We’ll get out of AI whatever we put into it. There’s this old acronym I learned a long time ago: GIGO. Garbage In – Garbage Out. Whatever amount of time, effort, and care we’re willing to put into it will go a long way to determining what we’ll get out of it.
3. Democratizing Discipleship: Lowering the Barrier, Raising the Depth: I don’t know how many of you are like me. Memorization is hard. It’s like when I was working – I had dozens and dozens and dozens of computer manuals. They were constantly updated. Or replaced when new systems came in. I couldn’t begin to remember everything. But I could remember how they worked – and from that could easily determine which manual I needed to look up details.
   With Bible Study, memorizing is super important to some people. But for those of us who just can’t – remembering a few words – or now with AI, even a concept pr a scene from the Bible – that’s plenty to find what you’re looking for. And it’s so much faster than skimming through chapter after chapter, book after book, trying to find something.
4. Structuring the “Tech-Shed Guide” Section: – regardless of the title – AI can save so much time. Given a whole bunch of information, it is usually very good at giving an outline, flow, structure, whatever you’re looking for in that regard.
5. Connecting it to Your Thesis: Gemini’s reply says it quite nicely: By focusing on the “teaching them to obey” part of the Great Commission, you elevate AI from a parlor trick to an act of stewardship. You aren’t using AI to replace the Holy Spirit; you are using it to replace the index of a book so you can get to the heart of ministry faster.
   I count on the Holy Spirit to guide me as to what to write about – points to include, based on the research, and especially for my comments that go along with it all. But in the end, AI is a tool to help me do those things, not to do them for me. The difference is having God’s love – or not. It’s following the Holy Spirit – or a machine.

---

At this point, I was done. But Gemini always wants to keep the conversation going. This time, the question seemed helpful in getting one last set of thoughts in for how I use AI.

My response:

You included early on that AI can provide structure. That’s one of the things I very much appreciate. Too many thoughts, too many details, and such. I look back at my writing pre-AI and much of it was, as one AI told me, pretty much a train of thought brain dump. But if I take the time to enter what I want to do – like this – and be sure we’re on the same page, an outline can be produced – you provide some details – and then I can expand on that with more research, excerpts, passages, and my own thoughts and explanations. So – like now I have something to run with – start writing, probably more research/questions – and then get all the technical things to publish on the web: SEO meta, keywords, tags, etc, so much more quickly than if I had to think of them.

And Gemini’s reply back to me:

It sounds like you have identified the ultimate sweet spot for human-AI collaboration. You are using AI as an exoskeleton for your intellect—it provides the rigid structural framework so that your unique human insights, research, and voice can fill it out without collapsing into a “brain dump.”

By aligning upfront, you ensure the outline matches your vision, allowing you to focus your energy on the actual writing. Then, handing off the mechanical, tedious tasks like SEO meta descriptions, tags, and keywords to the AI saves you hours of friction on the back end.

You have a highly efficient, repeatable workflow ready to go for this piece.

🗺️ Your Article Roadmap

As you transition into your writing shed, here is the structural skeleton we built today that you can run with:

• The Hook: The Barna trust gap (12% pastors vs. 48% practicing Christians).
• The Problem: The illusion of neutral synthesis and the critical need to identify hidden training biases (e.g., the Calvinist dataset example).
• The Reframe: Shifting discernment from “who said it” (source) to “what it produces” (fruit).
• The Core: The Tech-Shed Guide for Modern Disciples—using semantic concept search to fuel the “teaching them to obey” half of the Great Commission.
• The Practical: Multi-model triangulation and forcing the AI to show its historical/theological library.
• The Finish: The technical handoff (SEO, keywords, formatting) to get the piece web-ready.

---

Go ahead and start drafting, expanding, and weaving in your personal excerpts. When you are ready to come back to the tech-shed, I’ll be here.

Whenever you want to take the next step, let me know if you want to:

• Review your rough paragraphs to sharpen the focus and clean up any lingering “train of thought” tangents.
• Generate your SEO assets (meta descriptions, focus keywords, alt text, and tags) once your draft is finalized.
• Deep-dive into a specific safety check or find counter-arguments to include for a more robust article.

🤣 A final note on tech-shed 🤣

As mentioned, I didn’t know what Gemini meant by tech-shed. I also didn’t want to interrupt the session by asking. Sometimes they get easily sidetracked, not really being able to tell when something is just a side question.

So – later on I asked Copilot if it could tell me what tech-shed might mean. Its answer was very plausible.

After asking it to review this entire session notes + comments, my question to Copilot was: I’m curious – Gemini came up with tech‑shed – do you have any idea what it was supposed to mean? I didn’t want to interrupt the chain and ask at the time.

Note the assured confidence on an answer to what a different AI processed and came up with in the response below.

I had to respond to that: Actually – Now I truly like it. It’s a real/live example of what can happen – and it’s both funny and useful.

Once again, Copilot’s answer was interesting.

As I said – interesting.

---

Conclusions: How Christians Can Use AI Safely

Can Christians really use AI – safely? Yes – as long as, with everything in life, we’re careful.

Let’s be honest. Some churches claiming to be Christian don’t follow what Jesus taught in the New Testament. Christian Nationalism isn’t in the New Testament.

The Bible is full of warnings, even to those in the early church in New Testament times to be wary of false teachers and false Apostles.

Over time, methods change. But still, from the beginning in Eden, Satan’s goal has been to take us away from God. Whether it’s spoken words, printed text, computer generated text & images, the same warnings apply.

What I tried to provide here is one way – the way I use – to navigate AI in a manner where we’re aware of the potential for being led astray, but also doing what we can to not be taken in. It’s based on my decades of experience in IT, and more recently my 15+ years of doing Biblical research, study, writing and teaching.

Is it the only way? No, of course not.

Is it foolproof? No – nothing in this world is. Ever since computers were attached to networks, the only safe computer is one that’s not plugged in / turned on. But now, there’s some who would question whether even that’s safe enough.

Should Christians avoid AI? Truthfully, some should. If you don’t want to take the time to learn how to be safe with it, there are other ways, the “old-fashioned” ones, to learn about Christianity.

On the other hand, if you’re willing to take the time to learn how to use AI safely, it can open up a whole new way of learning, scouring resources to an extent never before possible. It’s amazing.

I remember a line about “swimming is staying alive while you’re in the water”. There’s a reason I rarely go in water that’s too deep for me to walk in. Those few occasions I’ve been in deeper water were to go snorkeling. I spend the first 15-20 minutes calming down and getting to a point where I can breathe smoothly through the tube. Then – after all that – I can enjoy the view. Why? ‘Cause I can’t swim.

But I’m OK with that. It’s just the way it is. We don’t have to do everything.

So, if you don’t feel comfortable using AI – it’s OK. People have become Christians for almost 2,000 years without AI. AI isn’t an essential part of knowing God well enough to be saved. Besides – in the next life, there’s an eternity to learn about God.

I use it because with my background, I can. And since I feel called to write and now to teach, it helps me go to another level with both. But lots of people do just fine without it.

Ultimately, AI is a tool. It’s not a destination or a requirement. But if you do choose to use AI, be safe doing it.