Using the Persona Approach with AI

Let’s dive in!

Many people hesitate to use AI, often because they haven’t explored its full potential beyond simple search queries. Rather than relying on secondhand opinions, the best way to understand AI is to experiment with it yourself. In this post (and upcoming ones), I’ll introduce practical ways to engage with AI, helping you move beyond skepticism or uncertainty into hands-on exploration. My goal isn’t to convince you to use AI—it’s to give you the tools to make an informed decision for yourself. Ready to begin?

AI Works Best as a Thinking Partner, Not as a Replacement for Thinking

AI should not be used as a replacement for thinking. Truly, it is your own creativity that is key to unlocking its usefulness. It is a wonderful thinking partner. What’s so wonderful about LLMs is that you can simply write out what you want it to do and it will work under the surface to write its own programs to accomplish what you need. If you don’t get the result you want, tell it what to change and it will try.

In the past, you would need to learn how programming works, write custom code that was highly specific to particular kinds of data, and a lot of other cumbersome steps. Now you can know literally nothing about the stuff under the surface and still come up with sophisticated help for an unlimited variety of things.

What is the Persona Approach?

The persona approach is where you ask a Large Language Model (LLM) like ChatGPT to take on a persona as it responds to you. By default, ChatGPT’s knowledge base is as big as the internet. Improving the clarity of your prompts and increasing your specificity helps the LLM know which parts of its information to pull from in helping you. Rather than you having to find spreadsheets of specific data, or work at length to describe a particular personality type or skill set, you can simply tell your LLM to function as one of these things, and it will automatically focus itself on the knowledge set that is most useful to your needs.

What’s the task you’re needing to accomplish? Wouldn’t it be great to have an expert in that field to guide you? Tell ChatGPT to become the expert you need. Here are some prompts you could enter into the LLM. Simply copy and paste any of these in to ChatGPT or another LLM and see what happens. If it asks you for more information, give it what it needs to proceed, then see what happens.

• I am working on planning a retreat for 7th-9th grade boys and girls for my church. I want you to think like a parent of a teen and help me work on what and how I need to communicate about the retreat plans.
• I have prepared a short devotional talk for elementary-aged children. I would like you to function as a 1st grader and have a look at what I’ve prepared, which I will copy and paste below. Then I want you to suggest how I can make it better.
• I am teaching a class about Exodus 3-4. I want you first to thoroughly familiarize yourself with all the contents of Exodus, then I want you to function as Moses himself and talk me through some of the highlights of the stories in chapters 3-4 from Moses’ perspective. What details might I easily miss? Tell me what it was like to be you.
• I am working on helping my young child learn to read. I want you to take Matthew 18:1-5 in an easy-to-read Bible translation and add lots of emojis to it so that as my child practices reading, the emoji pictures can give him some clues about the words he is trying to pronounce.

I have found it particularly helpful to invite AI to take on the persona either of something I’m studying or of the audience to whom I’m speaking. It has provided useful insights every time. For relevant idea generation and for relevant refinement, AI is an incredibly useful thinking partner.

➡ Level Up Your Study with AI Panel Discussions

I have debated keeping this one to myself because of how much I enjoy it, but I’m going to share one of my very favorite prompts with you that I’ve created. I started by asking AI to take on a specific persona here or there, but now in preparation to preach, I have started having ChatGPT generate a panel discussion each week to debate the meaning and application of the passage I’m going to preach or teach. Because AI is capable of doing deep research into figures for whom there is abundant material, it can simulate people effectively.

I always have at least one discussion round, and sometimes I go for two or three. On occasions, I decide that another figure could further enhance the discussion, so I tell it to add them. For example, one week I had C.S. Lewis contributing to a panel discussion and I decided to have Screwtape the demon show up and argue with him. Other times I’ve had Darth Vader join the panel.

I have always loved sermon research and preparation but I am truly having the most fun I’ve ever had in preparing to teach. The best part is that this highly entertaining method actually does lead me to useful insights and perspectives that shape how I prepare my content.

A great truth I’ve learned is that any time there is both a fun way and a tedious way to do something important, always pick the fun way! You will accomplish much more by being playful than you will by being stern with yourself.

➡ Sample AI-Generated Panel Discussion: Exodus 3-4

Here’s a sample discussion snippet based on the panel discussion prompt. This example assumes a panel discussing Exodus 3-4 (Moses at the burning bush) and features:

• Bertrand Russell (atheist philosopher, antagonist)
• C.S. Lewis (modern Christian thinker)
• Gregory of Nyssa (ancient Church Father)
• Jordan Peterson (public intellectual)
• Teresa of Ávila (Christian mystic)
• Abraham Lincoln (historic leader)

---

Panel Discussion on Exodus 3-4: Moses and the Call of God

Moderator:

Welcome, everyone. Today, we’re discussing Exodus 3-4—Moses’ encounter with the burning bush and God’s call to lead Israel out of Egypt. Let’s begin with Bertrand Russell, who has been critical of religious narratives. Bertrand, what’s your take?

Bertrand Russell (Antagonist):

Frankly, this passage illustrates the irrationality of religious belief. Moses is alone in the wilderness, likely suffering from hunger or delusion, and he claims to hear a divine voice from a burning shrub? If someone made such a claim today, we would question their sanity, not build a faith upon it. The entire notion of God “choosing” people and intervening in history is a relic of an unscientific age.

C.S. Lewis (Christian Thinker):

I appreciate your skepticism, Bertrand, but you overlook a key point: Moses initially resists his calling. If this were pure fabrication, we’d expect Moses to leap at the opportunity for power. Instead, he argues with God, protests his own inadequacy, and begs for someone else to go. This reluctance speaks to the story’s authenticity—it reflects the real, lived experience of those called to something greater than themselves.

Gregory of Nyssa (Church Father):

Indeed! And what Lewis says ties into the broader Christian understanding of divine calling. God chooses the weak to accomplish His purposes. Moses, a stammering shepherd, is tasked with confronting Pharaoh. This is a divine inversion—God does not seek the strong and proud but rather empowers the humble. His “I AM” statement reveals His eternal nature—self-existent, beyond time, and beyond human comprehension.

Jordan Peterson (Public Intellectual):

Gregory, you highlight something profound. From a psychological perspective, Moses represents the individual confronted with responsibility. The burning bush is not just a miracle; it’s an archetype of transformation. The moment a person confronts their deeper purpose, they stand at the threshold of chaos and order—just as Moses does. He can refuse the call and stay in the wilderness, or he can step into the unknown and reshape history.

Teresa of Ávila (Christian Mystic):

Moses’ reluctance also mirrors the interior struggles of the soul. When God calls, we often resist, feeling unworthy. Yet divine love is persistent. Moses receives not just a task but divine companionship—God reassures him, provides miracles, and even grants Aaron to help. This is how God works in all of us: He calls, we resist, but He patiently leads us forward.

Abraham Lincoln (Historic Leader):

That idea resonates deeply. Leadership, particularly in moments of great upheaval, often falls upon those who feel least prepared. When I was elected in a time of national crisis, I felt an overwhelming burden. Yet, as with Moses, I found that duty is not about personal capability—it’s about answering a higher call. The Lord equips those He calls.

Moderator:

Fascinating insights! We’ve moved from skepticism to faith, from psychology to leadership. Now, let’s discuss a pressing question: What does this passage teach us about doubt? Is doubt a weakness, or is it part of faith? Let’s start with Teresa…

---

Why This Works

• Dynamic Interaction: Each panelist contributes something unique and builds off the previous statements.
• Diverse Perspectives: The discussion covers philosophy, theology, psychology, leadership, and mysticism—making it richer than a single perspective.
• Engagement with the Text: Rather than just debating belief vs. atheism, the conversation explores the nuances of Moses’ resistance and God’s response.
• Usability: You could adapt this for any biblical text by selecting new panelists and refocusing the discussion.

---

Part of how I’m trying to teach you to utilize AI is by demonstrating best prompt practices. Read through my prompt below first so that you can understand my thought process and the way in which I’m guiding the LLM to work with me. Next, copy and paste the prompt below exactly as it is into an LLM like ChatGPT. If you like the discussion round one, have it go for as many rounds as you like. Trade out the panelists. Have them become more or less argumentative. Insert fictional characters or aliens. It can do whatever you can think up!

I invite you to try this out and see if you don’t have a blast with it:

I want you to produce a panel discussion that will involve interactions between several interesting people. I want you to ask me only one question at a time if you have any, but before you generate the panel discussion, first ask me about the passage or topic I want the panel to discuss and also provide the option of you suggesting a topic.
After I have decided what the panel will be discussing, I want you to select people from the following categories to participate in the panel.

1. A historical person of significance who has been antagonistic to the Christian faith.
2. A great modern Christian thinker.
3. An ancient Church Father.
4. A public intellectual.
5. A Christian Mystic from a previous century.
6. A historic national leader or president.

Select panelists who make sense and have special relevance to the topic. If you have run this prompt for me previously, select different people than you did for the previous attempt. After you have selected them, present me with your list of suggested panelists with a brief note about who they are and why they could contribute to this discussion. Ask me if I am satisfied with this list or if I would prefer any changes before you proceed.

After I indicate I am satisfied, take a moment to familiarize yourself thoroughly with the panelists. Learn all about their works, their philosophy, their theology, and their style. Be able to speak competently as each of them. After you’ve done this, begin generating the panel discussion.

Begin with the antagonistic person who makes an attack on the value of this passage or topic, particularly an attack on the Christian perspective on the topic. Each of the participants should say 150-200 words when they respond. If they would be inclined to affirm or deny a point from another panelist, have them interact with each other.

After the antagonist has spoken, then one by one have the others respond to the first and point out holes in the antagonist’s arguments. After they’ve responded to the antagonist, have each of them follow up by expressing what they see as the valuable possibilities for how to use the passage.

You function as a moderator and at the end of the conversation cycle, summarize the takeaway points from the conversation. Ask me if I am satisfied, or if I would like another round of discussion.

If I decide I want another round, first ask me if there is anything I want to change before you proceed, such as an additional panelist or a more specific topic I want the panelists to debate. Then, based on my input, proceed with the next round of discussion.

In my experience, bringing people from different disciplines and different centuries together to debate the topic is an incredible way to add depth to your preparation and fresh perspectives that I wouldn’t have been able to generate from within myself.

Also, as you probably noticed, this is a large prompt. Remember, the more specificity that you provide to a large language model, the better quality response it can give you. A highly specific prompt can produce an incredible result.

Consider, also, that you could tweak this approach for an unlimited number of uses:

• A multi-generational panel to discuss your sermon ideas with you
• A panel of leaders from different denominations or different world religions to discuss controversial topics
• A plumber, a carpenter, an electrician, an architect, and a designer to talk about a space you are thinking of changing or enhancing in your building

Would any of this be the final word on what you end up doing? NO! But could it get you off to a great start and help you steward your time and resources? Absolutely!

Important to Note: AI Creates Text, Not Truth

One thing that it is important to understand about generative AI is that it is designed to generate text, not information. It tries hard to produce the kind of thing it understands that you are expecting. One good practice you can notice above in my larger prompt is that I tell it very directly to do good research prior to generating a response. I urge it to be thoroughly familiar with the words, theology, and philosophy of all panelists. Those guardrails are important.

In the free version of ChatGPT, my understanding is that its knowledge base stopped in 2021. The newer version, as well as programs like Perplexity AI, can utilize up-to-date web searches for content. But it really matters what sources of information AI is pulling from as it response to you. Remember:

• AI doesn’t “know” facts—it generates plausible text based on patterns.
• It can hallucinate information, especially if you don’t set guardrails.

Sometimes AI will give you the kind of text it thinks you are wanting, but is totally making up the information. For example, the very first time I used generative AI, out of curiosity, I asked it to write a sermon about a topic, wondering what it would produce. It was a well-organized outline. Then as I began looking at the Scriptures it was quoting, I was impressed there were some I wasn’t familiar with. Then I began looking up Scripture references that it quoted, only to discover that they were totally made up. No wonder I was unfamiliar!

When AI generates nonsense or made up information, people call this hallucinating. As I said in my previous post, I do not and do not recommend asking AI to do the work that you yourself should do in researching and speaking from your head and heart. This is a great example of how a person could embarrass himself or herself if you simply believe whatever AI tells you is accurate and run with it. Always verify!

If you ask AI how your doctor visit went last week, it might try to tell you! And it would be totally wrong! This is because it’s simply trying to give you text that it thinks you’re asking for. If you don’t provide the guard rails about the importance of fact verification or quoting sources for certain purposes, it might just make up whatever it needs to provide the response you requested.

How to Avoid AI Hallucinations:
✅ Always verify scripture references AI gives you.
✅ If using AI for research, ask for its sources and then double-check the sources.
✅ Be careful: AI creates content—it doesn’t necessarily confirm facts.

An Asset, Not a Problem

However, the ability to imagine content is actually an asset and not a problem with AI. If you understand that AI is simply trying to write with text what it thinks you are wanting, and NOT that it is a source of all information like Google, you can use it to do amazing things, such as the simulated discussion between great thinkers! I would never quote those panel discussions, because I know as I’m reading it that AI could be wrong about the details and misrepresenting what the panelist might actually think. However, the characterized kinds of responses it creates are still useful perspectives. The antagonistic perspective, in particular, I find highly useful in thinking about preaching and teaching. What might a skeptic say? The AI generation will be a good guess!

How are you using AI?

I know my ideas are just a drop in a vast, vast bucket of amazing possibilities. Are you using AI? What are you using it to do? What prompts have you found especially useful?

I have much more content to come on this topic. If you haven’t already, I hope you’ll subscribe to Kingdom Upgrowth so that you can be the first to hear about new content as I post it.

Other Content You Might Enjoy: