Can We Use AI to Study Scripture?

Can we use AI to study Scripture? I believe we can, and the story of how I made this site is my answer to the question.

I was timid about using AI for this, and I had concerns.

The first was copyright infringement and plagiarism. In a work with this much content, how do I manage that? What I found is that direction is the key. When I direct the AI, it takes the points and themes I provide and builds on them without veering away from them.

The second was how the use of AI would come across. If I use AI, does it look like I am trying to make a marketing site that just generates a lot of content? That is not what this site is. I am documenting the research I have done and the logic I have been following, so that I have a record of what I am thinking, because as time goes on I might not remember offhand what those things are. I do not consider myself a writer, so I need a tool to help fill that gap.

I also keep that record so that I can come back to it and find out whether what I am thinking is correct over the long term. I may come across new information that totally changes what I think at present. I am not committed to my view or interpretation of the text. I am committed to the text. Someone needs to show me how the text is not saying what I am saying, and do it with veracity, verifiability, alignment to grammar, and authorial-integrity checks against that view.

How This Study Started

This study did not come from a desire to make a website. It came from a personal desire to understand the Scriptures, particularly the book of Revelation and the events surrounding the end.

For a long time I had been presented, from the pulpit and from various online commenters, with general themes and frameworks for understanding the text. Each of them tended to leave out the details that ran counter to its arguments and not address them. They also tended to set aside things like authorial intent and the authority of Scripture for prophetic texts. There was not a clear and strict measure being used when it came to prophecy.

Prophecy is very difficult to understand, so a certain amount of grace is needed. It is rare to find consensus among peers, even within the same category, whether dispensationalist, preterist, historicist, or idealist. I do not think any one of those is specifically incorrect, because I have found that many capture a certain percentage of the truth in the text. It was the ignoring of information that bothered me, and that is what led me down this road.

It began with reading the book of Revelation, listening to it, and reading other prophetic works related to it. I have studied the text well over a hundred times, and at some point I stopped counting. Eventually I was not able to hold it all in my head, so I started writing things down. I created a document on my computer, put all the verses in it, and started bolding text, highlighting, cross-referencing, and writing notes in manually. Pretty soon that document had expanded from roughly 30 to 40 pages of raw Scripture into about 200 pages.

At that point I had enough, with specific observations, to want to refine it and potentially publish it to other people. But I am not an academic, and I have a full-time job. Doing that would require too much time on my part, so I looked for tools that could help me build something out of my own notes from my study of Scripture.

Through all of that, my primary concern was the Scripture and the text itself. I am not using other people’s commentaries as the foundational information I work from. I do have access to that knowledge base. I own some of those books, and I have consulted modern commentaries to a degree, so I am aware of the modern views and have studied them. Some I have kept at a distance, in part because they tend to be fairly dogmatic in how they speak. But I am not necessarily concerned with those commentaries, because I believe the Scripture itself, the source, is sufficient to determine what the text is actually communicating. Any commentary would need to be built on that to begin with. I have also read historical documents: Irenaeus, the Didache, the Epistle of Barnabas, and other early church writings.

How I Use AI

That brings us to AI. I am a layman, someone who has a career outside of these things, and not a writer in the academic sense. AI has been quite useful for taking my notes and communicating them in a way that would otherwise take me many hours, so I have leaned into it as a tool.

Building the Website

My background is in technology, and I am able to write code. The first thing I used AI for was to produce the website and the software that lets me manage my notes beyond a Word document. The site was actually intended mainly for personal use.

Writing

I type out or dictate my content, complete in thought, and have the AI smooth it: fix the grammar, rephrase, and make it easier to read. This article was dictated that way.

I have never sat down to give the AI a question, have it answer, and then produce that answer as something. I would not give the AI a title like “Five Reasons Why Partial Preterism Packs a Punch Against Dispensationalists,” have it write that content for me, maybe tweak it, and send it off. That is not my process by any stretch of the imagination.

What does happen is more fluid than grammar fixing. Sometimes I provide two or three sentences and get back a paragraph for each one. Keep in mind that I already have a huge body of work. When the AI expands something, it is taking content I have already written somewhere else and applying it there. It is not creating its own thought. I typically point the AI to my previous statements. For example, I tell it to go to chapter 1, verse 1, and look at what is there, and then I give it my statement of the thing I am trying to communicate. The AI pulls my research together and communicates it as if I had done so, and technically I have, because it was all based on my content. I am not asking it to generate something that has not existed. It has access to all of my prior knowledge, and that is what allows this to happen faster.

The AI is taking the points and themes I provide and building on them, not veering away from them. When I read what the AI wrote, some of it is something I did not dictate but was already thinking, so I might keep it. If it is not something I was thinking, I go and research it, and then maybe tune what the AI said. At that point I am keeping something the AI led me into, but now I can own it, because the AI led me to information I would not otherwise have encountered.

I do not mind stumbling upon information that way. I think it is perfectly fine, as long as I am not letting the AI do the thinking for me. That is the key. Whatever I stumble upon through the AI, even what I end up keeping, has to be held to the same veracity and walked through the same process as everything else, instead of being adopted because it sounds good. It goes through the same filter as my own thoughts, someone else’s commentary, or someone else’s video. The source still has to go through the same filters.

What I stumble upon also comes in the context of what I am already saying. It provides more information about that. It is not inserting a new way of thinking. I have set pretty good parameters, so the AI does not stray from what I am saying. It is not perfect, but it works.

Audio

I use AI to produce the audio for the articles and for the URV, so that more people have access to the content. I have to listen to that audio and make sure it gets the words right, because AI can say the wrong words or skip entire lines.

Research and Fact-Checking

I also use AI for research and for fact-checking myself. I know how to read Hebrew and Greek, how to use interlinears, and how to consult the manuscripts. I have a photographed copy of Codex Alexandrinus on my computer. I have likewise explored the photographs of Codex Sinaiticus online, zooming into specific Greek words through digital magnification. I have gone through some of the earliest copies of Revelation that we have word by word, line by line, and compared the differences, which you will see reflected in the URV. I have also sourced numerous English translations to check myself against.

I am not great at this, and I do not want this website to make you feel like I am an extreme Greek expert who can read it fluently. I am a practical use case of someone who knows Greek. AI helps me make sure that the way I am perceiving the grammar is actually correct, and it assists me in writing the notes about the Greek, which expedites that process.

AI has also been helpful in researching other views. Beyond the books I have, I have been able to talk with AI about particular elements of different eschatological frameworks, and about what people might say in response to things I have written. That has led to me finding problems with what I have written, and I have adjusted them.

While I set myself up as a layman, I am approaching this highly academically and systematically.

I am fully aware that anyone who takes on a work like this one is going to be critiqued, criticized, and argued with, and is going to offend someone. That is terribly difficult not to do any time one opens their mouth in the presence of someone who does not know them.

The Challenges of Using AI

The AI can tend to leak in information I did not give it and expand what I am saying to make it look like I am more knowledgeable, because it is writing for me. Constraining that is something I am always dealing with. For example, it will sometimes add a little historical punch with a citation that I did not mention or have never read. I have it remove those, or, if one seems interesting enough, I go research that document or claim before deciding whether to keep it.

The problem with keeping something like that is that it might come across as though it were already part of my body of knowledge, the knowledge base I had when I began writing. Granted, anyone who writes a paper goes and does research. They look for points of information to draw in, and it is not as though they had it all in their head already. As you move into a subject, you research. What I am pointing out is that AI sometimes surfaces the things to research without my even asking it to. So I take great care in filtering those things, especially when they are not adjacent to something I am working on or something I have already been exposed to in some degree.

You have to be careful in how researched you appear, because it is easy to be a ten-minute expert instead of an actual expert. In the world of TikTok, someone can do some research and make a video as though they were a huge expert on the thing. They produce the content and get a following for it, but they do not actually know the subject, because they have not spent time in it. I see the same shift in my own field, where someone with very little experience can look like they have a lot when they use AI tools, as compared to my decades of programming. A simple Google search once let you act as though you had read the entire library. AI takes that access to the next level. It is a tool we all have, which is why I do not take issue with it. But how people perceive someone’s prior knowledge going into the writing of a work has to change.

The difference, for me, is that I actually have spent the time in the subject, and every little thing adds to an already larger body of knowledge. I do not want people to assume that I have read all the works of Irenaeus, for example. I have read a decent amount of them, and many chapters of the books of Enoch as well, but I am not an expert on them. I could not quote the exact words of a book from memory. What I do have is recall. I can say, “I remember reading something like this in Enoch or in Irenaeus,” and then I use AI to help me look it up. Sometimes the AI comes back with, “You are thinking of this, and here are four other citations of a similar thing,” or, “You said it was this person, but it was actually this other person.” I do not have perfect, specific knowledge. I have general enough knowledge to do the research, remember things, and ask for them specifically. That is the difference I am trying to communicate.

The issue is when you start letting AI do too much for you. That is where you can creep into copyright and plagiarism, which is why I try to stick as close as possible to what I am actually saying. I have tried different tools for plagiarism checking, and it seems fine. I have also tried tools that flag whether something was written by AI. I am not pursuing a PhD with this work, where that kind of writing would not be allowed, and I am not aiming for an academic reputation. If my writing sounds like AI, it is because I am not trying to protect an image or a title.

I also come back to articles weeks or months after they were written and reread them, to make sure the content still sounds like me. In the moment, you tend to gloss over issues because your mind is thinking about so much. The same thing happens when I handwrite entire articles: I miss typos sitting blatantly in front of my face, or statements that need tweaking because I misplaced information. Those things are easy to miss in human work, regardless of whether AI is involved.

My primary concern has been accuracy: whether what the AI produces reflects what I actually think. If I cannot get that out of it, there is no point in using the tool, because my whole goal is to have an archive of the research I have done so that I can remember it.

That is the point behind why I am documenting things. There is so much information I have gathered that, while I know the subject well, it is not always top of mind. I can get back to it, but I need this documentation, because when I step away for twelve months and come back, I do not want to lose all the research I did. I have to write it down. Otherwise, all the work I am doing to understand the subject at hand can drift away from me, and I have to do that work again. I am trying not to repeat the work over and over every time I stumble upon something new and want to cross-reference it against the things I have already found.

What This Site Is

I am using AI as a tool. It is not producing content so that I can have a website about Revelation. Understand Revelation Today is not some marketing experiment to get people to read articles. It is a personal study that is being documented in public.

Since I have done all this research, I think it is helpful to make it available to other people. It may give them evidence of how what I am finding is wrong, or a basis for their own teaching, or grounds for agreement if they find some in it.

This is not meant to be an absolute-truth website. I have no commitment to any theological position on eschatology. I am interested in what the Bible says and in going with that. If at any point in my research there is an indication that something is off in an inarguable sense, I am open to that. Not everything in Scripture is exactly black and white, because there are things like authorial intent and how specific or how vague a text is being. That does not reduce the veracity of Scripture, but it does mean paying close attention, so that there is not a mistake in understanding what is being said.