Editor’s Note: Is generative artificial intelligence (GenAI) a morally neutral tool helpful for generating sermon outlines and small group discussions? Or is it something attractive on the surface but deceptive in its core—which we should avoid in our ministries? As with other issues not directly discussed in Scripture, navigating GenAI calls for a lot of wisdom. Doug Smith is a decades-long software engineer, a passionate student of the Bible, and an award-winning author focused on helping families survive a screen-saturated world. As a tech insider, Doug’s guidance is valuable as we each cultivate our convictions regarding GenAI.
Surprised By Magic
It’s hard to remember that ChatGPT first enchanted the world back in November, 2022, only about 2.5 years ago at this writing. Since then, Big Tech’s top priority has been to captivate the world with their Large Language Model Generative Artificial Intelligence chatbots (GenAI).
Maybe you, as a Christian leader or disciple, have also been using GenAI to improve your efficiency. Maybe you’ve been summarizing meetings, brainstorming, or otherwise generating words, code, images, and videos. Maybe you’ve been impressed with the power, but also confused by the sometimes nonsensical output and fabricated references.
If you’ve been using GenAI for a while, you may have started to sense a dependency forming—if a question enters your mind, a chatbot is the first thing to which you turn for answers.
But you’ve told yourself that it’s a tool, and as long as you’re careful and just use it for good, and maybe take breaks sometimes, you’ll reap the benefits without consequences.
If that’s you, you’re not alone.
“You’ve told yourself that it’s a tool, and as long as you’re careful and just use it for good, and maybe take breaks sometimes, you’ll reap the benefits without consequences.”
Embedded in Big Ministry
Craig Groeschel, pastor of LifeChurch.tv, maker of the mega-popular YouVersion Bible app, is pastor of one of the largest multisite congregations in the world. You’d be hard-pressed to find someone who speaks to more Christians on any given Sunday.
On a recent Carey Nieuwhof Leadership Podcast (February 18, 2025), Groeschel shared how he’s using GenAI in his ministry. He also confessed his discomfort. He believes we can use it wisely, but also said, “I think that AI is going to be both a gift and a curse to us.”
The five-minute segment brimmed with the tension Groeschel feels. He said, “I’m both optimistic and afraid of it. I’m optimistic because I think it can be helpful, but when it starts to replace our own private study, devotion, creativity, I think it’s going to be very counterproductive.”
But as he described his usage, it seemed to me that GenAI has already become Groeschel’s intimate ministry partner. Consider what else he said:
- “I will ask it dozens of questions every day.”
- “I’ll put in a message [sermon] and say, ‘Critique this. Where is it weak?’ And it’s surprisingly good, surprisingly good at times.”
- “I’ll ask it to tell me what people would say in their minds about this subject.”
Take a step back and put almost anything in the place of “it” in the above quotes. Dozens of questions every day. Ideas to shape the teaching of the Word of God for the people of God. Inspiration. Collaboration. Critiquing what he says to his congregation.
We ask questions of people we trust. By definition, when we ask a question, we’re hoping for help, and we bestow at least some authority on the person or thing we’re querying.
“We ask questions of people we trust.”
I don’t know how else to say this: GenAI has become an influential, trusted partner of the biggest megachurch pastor in the country. Groeschel described an intimate, personal relationship with a chatbot at the heart of his ministry.
After confessing all this usage to Nieuwhof’s audience, Groeschel admits his fear:
“Sometimes I get so stuck in it, I say, ‘No more, I’m done. I’m not going to look at it again today because I don’t want to become over-reliant on it.’ So then I put it aside and force myself not to use it anymore. So it just needs to be a tool, but not something we want to rely on solely.”
Groeschel senses he’s getting too dependent. He doesn’t want to be over-reliant. He thinks he can use it in moderation, wisely, as a “tool.”
So many of us do the same thing: we excuse our use of many technologies, including GenAI, by telling ourselves that it’s just a tool and it’s how it’s used that matters. This happens so often that I call it the “tool trope.” And it is simply false. I’ve made that case here and here, and I’ll explain it more below too.
But before we consider whether we can use GenAI merely as a “tool” in wise moderation, we need to understand what today’s GenAI really is. And before we can understand the artificial, let’s take a quick look at the real knowledge that fuels true intelligence.
“Before we can understand the artificial, let’s take a quick look at the real knowledge that fuels true intelligence.”
How Do We Know?
In his landmark book on epistemology, Personal Knowledge, Michael Polanyi taught us that the ability to know something requires three capabilities[1]:
- the ability to perceive reality,
- the ability to compare reality to an abstract concept under consideration,
- and the ability to choose the right words to describe the concept based on reality.
God has given his image-bearers this gift of knowing, and of putting words to what we know.
For example, classical musicians know what it means to play staccato. We know that because our teacher played short, punctuated notes, and told us that was staccato. Then she taught us to play staccato ourselves. We experienced the reality of staccato, envisioned the concept of staccato, and then learned the right Italian word.
In order to be intelligent, to know something, we need these capabilities. And—spoiler alert—GenAI has none of them.
“In order to be intelligent, to know something, we need these capabilities. And—spoiler alert—GenAI has none of them.”
What is GenAI, Really?
Few of us are data scientists. But even those professionals don’t always see why GenAI chatbots are so captivating to us, and why they are dominating every sphere of life, including the pulpit.
My picture of GenAI’s captivating power includes four circles in a stacked Venn. The core is the innermost circle, and the last is the outermost that surrounds the rest.
1. GenAI’s Statistical LLM Core
At the core of every GenAI chatbot is the large language model (LLM) itself: gazillions of “parameters” which define the statistical relationships between words that were created by training on whatever input data the company sourced. GenAI is sometimes rightly described as “autocomplete on steroids,” choosing the next word by passing the previous words in the conversation through a mind-blowing calculus.
It’s vital to understand that at the core, GenAI chatbots understand no abstract concepts nor do they experience reality. They have no context nor connection to the real world. That’s why chatbots can be so wrong. It’s not just that they “hallucinate” a lot, but that every word is statistically chosen and therefore ungrounded.
So when a chatbot strings a series of words together that sound convincing, know that those words were not chosen by understanding a concept or by experiencing reality, but only through statistics. They were made up, with no connection to understanding truth. As Christian leaders, we should care about that.
“Those words were not chosen by understanding a concept or by experiencing reality, but only through statistics.”
2. The Natural Language Processing User Interface (NLP UI)
All modern chatbots are wrapped in a powerful natural language processing user interface (NLP UI). They are intentionally designed to feel intelligent, reasonable, wise, discerning, and to emotionally connect. The conversational user experience is not an accident. It is designed to show confidence, but also empathy, and even to apologize when called out for inaccuracy. To act like a person. And especially, to foster an intimate relationship.
Even though the statistical core has no grounding in reality, the NLP UI wrapper is intentionally designed with a confident and emotionally compelling tone to make us feel like there’s intelligence and wisdom and knowledge there. But there’s none.
You know why the UI works so well on us? Because up until now, we’ve only chatted with other people—some of us for decades. We’ve used Slack or Teams or Messages or iMessage with our families and coworkers. And chatbots feel just like that—exploiting our prior experience, by design.
When I think of a system that pretends to be authoritative, wise, confident, caring, and even sentient, one word comes to mind: deception. We’re in danger of being deceived. And as a Bible student, I know where all deceptions come from: the father of lies (John 8:44). As Christian leaders, we should care about that.
“It is designed to show confidence, but also empathy, and even to apologize when called out for inaccuracy.”
3. Stories We’re Supposed To Believe About GenAI
Wrapped around the LLM core and the NLP UI are a host of stories that we hear from Big Tech, that we hear from other people, and that we tell ourselves. These stories inform what we believe about chatbots and entice us to accept them.
For example, a colleague who uses ChatGPT believed that it was doing “a million Google searches in the background.” He didn’t understand why I would claim that it makes stuff up until I showed him the statistical, ungrounded, inner core of the LLM.
In his conversation with Craig Groeschel, Carey Nieuwhof (who also seems all-in on GenAI) exclaimed, “It knows Hebrew? Of course it knows Hebrew!” This is a story. Chatbots know precisely nothing.
But we’re told stories that chatbots are “intelligent,” they “reason,” they “understand,” they “know,” and they have “context.” But none of these statements are true. They only feel true because of the first two circles, and we believe them because of the stories wrapped around them.
Finally, we’re given the yarns that prior technology revolutions were disruptive but ultimately good: the printing press, railroads, cars, computers. All of these things have helped us spread the gospel. So GenAI must be good too, right?
As Christian leaders, we should be much more discerning about the stories that justify our adoption of new tech.
“As Christian leaders, we should be much more discerning about the stories that justify our adoption of new tech.”
4. Stories Big Tech Leaders Believe About GenAI
Wrapped around the core, the user interface (UI), and the stories we believe, are the stories the makers themselves believe about what they’re doing. Those stories are driving them to risk hundreds of billions of dollars and to hype GenAI as the Next Big Thing™.
Sam Altman, CEO of OpenAI, maker of ChatGPT, believes he is creating God. Altman is a transhumanist who believes that humans will either “merge” with machines into some new species or we’ll become an evolutionary dead-end when AI takes the top species position.
Many other leaders like Dario Amodei, Ray Kurtzweil, Mark Andreessen, and even Elon Musk have utopian views of AI that form a techno-powered eschatology. Amodei predicts curse-overcoming health gains, proclaiming that “AI-accelerated biology will greatly expand what is possible: weight, physical appearance, reproduction, and other biological processes will be fully under people’s control.” And Andreessen shares a gospel: “I am here to bring the good news: AI will not destroy the world, and in fact may save it.”
These people lead some of the most powerful corporations in the world. They have global influence, so their propaganda fills the collective consciousness. They are the reason we tell ourselves such optimistic stories about AI, and even excuse our use of it with the tool trope.
Should we care about what the makers of today’s GenAI leaders believe? Can we separate their non-Christian worldview and intentions from their products and still use them for good?
“Can we separate their non-Christian worldview and intentions from their products and still use them for good?”
How Using GenAI Disciples Us
If we’re building a personal relationship with a product that is:
- a content generator with no grounding to reality that is often confidently wrong,
- wrapped in a natural language user interface (UI) that is intended to make us believe we’re talking with an intelligence when we’re not,
- wrapped in stories we tell ourselves about the power and capabilities and knowledge and utility of these systems,
- wrapped in the worldview of the creators that hopes to usher in techno-salvation,
then what does our relationship with GenAI do to us?
Iconic media analyst Marshall McLuhan taught us decades ago: every technology extends a human capability but then amputates the natural ability we used to have before we used the tool. And when tech is new, we are numbed to the amputation.
Through our ongoing relationship with a chatbot, we are transformed from an active creator (like God) into a passive consumer. In our consuming posture, our vital capacities of critical thinking, discernment, wisdom, and creativity are gradually amputated while we are numbed by the thrill of usage so we can’t sense what we’re losing.
“In our consuming posture, our vital capacities of critical thinking, discernment, wisdom, and creativity are gradually amputated while we are numbed by the thrill of usage so we can’t sense what we’re losing.”
The chatbot user interface (UI) exploits weaknesses in our behavioral psychology by applying the same dopamine-spiking techniques used by Big Tech to shape our desires via social media. We want to use and trust and build a relationship with chatbots because of their persuasive design.
The UI is also designed to foster what I call “incantational habits.” We are trained—you might even say discipled—into thinking that all problems should have quick answers, and that the essential skill is not to pray, study, and think, but to craft the right prompt (a.k.a. incantation) into a chatbot. We think that if we use GenAI “responsibly” we will retain our ability to discern, but we’re deceived. No Big Tech product works that way. Ask anyone who intended to be on social media or YouTube for five minutes and ended up doomscrolling or autoplaying for hours.
As we are discipled by GenAI in these ways, we may sense a change in our spirit. But like Groeschel, we excuse it with the tool trope, telling ourselves that we are becoming more efficient and that it’s helping us.
A Biblical Disciple’s Lens
Now that we see how GenAI is designed to shape our desires and amputate essential human characteristics, I want to ask some scripturally-informed questions.
If 1 John 5:19 (NASB) is right and “the whole world lies in the power of the evil one,” would we expect products made by the most powerful corporations in the world to be helpful to Christian ministry and witness? Helpful in crafting our messages?
If there is going to be such a “strong delusion” (2 Thessalonians 2:10–12, ESV) in the last days that include “great signs and wonders to deceive, if possible, even the elect” (Matthew 24:24, NIV), is it—even a little bit—possible that GenAI could be a tool to that end? Given that GenAI is a ubiquitous technology that deceptively pretends to be intelligent but is not, surprises us with magic answers to our incantations, and is habit forming while whittling away our discernment and critical thinking?
And if there is a “prince of the power of the air” (Ephesians 2:2, ESV) and our battle is truly “against spiritual forces of wickedness in the heavenly places” (Ephesians 6:12, NASB), but also that enemy can take the form of an “angel of light” (2 Corinthians 11:14, NASB), appearing attractive, useful, helpful, even as a blessing from God, might GenAI be a tool in that prince’s toolbelt rather than ours?
As I’ve said elsewhere, we dismiss questions like these to our peril.
“We dismiss questions like these to our peril.”
Did Jesus Command Efficient Discipleship?
Why are we so focused on efficiency? That’s the whole point of GenAI, right? The big selling feature: all of that pesky thinking and writing and brainstorming and summarization and critique and editing takes time and hard work. (Believe me, I know—writing an article like this is hard!)
Should we sacrifice these incredible God-given abilities on the altar of efficiency so we can reach more people?
I don’t think so. Because Jesus didn’t command efficient discipleship, nor did he model it. Regular RENEW.org readers know this well. Jesus’ way of discipleship was a rabbinic, all-in, 24-7, messy, authentic life with his disciples. Anything but efficient.
Yet the early church “turned the world upside down” (Acts 17:6)—not by the power of GenAI, but the power of the Holy Spirit.
My takeaway? Let’s do the same. Reject today’s hype-filled fad. The risks are too high. Instead, invite the Spirit to give us everything we need to lead us into all truth (John 16:13).
[1] Michael Polanyi, Personal Knowledge: Towards a Post-Critical Philosophy (Chicago: The University of Chicago Press, 2015), 95. Kindle Edition.
