Get Renew.org Weekly Emails

Want fresh teachings and disciple making content? Sign up to receive a weekly newsletters highlighting our resources and new content to help equip you in your disciple making journey. We’ll also send you emails with other equipping resources from time to time.

13 minutes
Download

5 Questions for Church Leaders to Ask About Generative AI

With the rise of generative AI, are we closer to Clippy, your old animated friend from Microsoft Office 97—or to the Terminator? Are we closer to Buzz Lightyear or to the AI-run jet from the movie Stealth?

Here are some snapshots of where we are: An AI-generated work took first place in a Colorado state fair’s digital category. We’re seeing articles called “I Coded a Website with 60k+ Daily Visitors Using ChatGPT” and “ChatGPT Could Make These Jobs Obsolete.” Artificial intelligence is flying F-16s.

The “Christian response” to such a speeding-train blur is predictably going to fork in multiple directions. You’ll see articles from “ChatGPT Has No Future in the Pulpit” to “Using ChatGPT in Sermon Preparation.”

In this article, let’s look at five questions at the intersection of generative AI (GAI) and church leadership. These are five questions I encourage church leaders to wrestle with regarding GAI:

  1. What has changed?
  2. How could we use it?
  3. Should we use it?
  4. What should we be cautious of?
  5. Should we feel threatened?

What Has Changed?

A couple ways to summarize human intelligence would be that 1) humans can decide between A or B, and 2) humans can create C. For decades now, humans have created and educated machines which are then able to decide between A or B. For example, our machines can discern the following types of information:

  • Is this the face that can unlock the phone?
  • Is this email spam or not?
  • Based on these weather trends, what will tomorrow’s temperature be?
  • Based on these purchases, what next purchase will you likely make?

More recently, humans have created machines which can “create C.” Here are options now available for basically anyone with a computer and internet (thanks to programs such as OpenAI):

  • You can type a description, and it can generate an image.
  • You can enter an image, and it can generate text.
  • You can enter picture and a text, and it can generate video.
  • You can type text, and it can generate essays and audio.
  • You can type text, and it can generate code for a website.

So, artificial intelligence is when machines can

  • Decide between A or B = “discriminative AI”
  • Create C = “generative AI”

For example, ChatGPT is an AI chatbot developed by OpenAI, an AI research lab funded by Microsoft. Launched in November 2022, ChatGPT has been trained on countless books and websites, especially Wikipedia. It’s also been finely tuned to generate human-like responses in a conversational manner. Thus, it’s able to distill all its voluminous information into paragraphs and essays, as well as combine its information in creative ways.


“It’s able to distill all its voluminous information into paragraphs and essays, as well as combine its information in creative ways.”


In 2023, one version (ChatGPT-4) was able to pass the bar exam, scoring in the 90th percentile. ChatGPT has since been integrated into Apple products. The OpenAI website has risen to be among the top 10 most-visited websites in the world. As of this article, GPT-4o is the latest model and integrates text, voice, and vision to have an even more authentic-feeling conversation as it computes, for example, your facial expressions and tone of voice.

How Could We Use It?

For content creators, there are at least a couple ways generative AI is being used that aren’t too controversial. First, it’s being used as a research assistant, and second, as a creative jumpstart.

As a research assistant, AI can make a traditional Google search seem clunky because there’s not the scrolling, clicking, and more scrolling that goes into letting Google guide you to the right source. It’s no wonder that Google increasingly saves its top spot for its own AI-generated answer. Unfortunately, its answers aren’t always accurate even if they have a ring of credibility.

Generative AI is also being used as a creative jumpstart. About a year ago, I was curious how “creative” ChatGPT could be, so I typed a prompt for ChatGPT which led to a surprisingly insightful answer. I asked, “What are three metaphors for forgiveness?” and it responded with 1) a balm that soothes a wound, 2) a key that unlocks a prison, and 3) a reset button for a malfunctioning system. Not bad.


“First, it’s being used as a research assistant, and second, as a creative jumpstart.”


Should We Use It?

Johannes Trithemius (1462-1516) was a brilliant German abbot who became one of the fathers of modern cryptography by inventing a cipher. Yet he strongly resisted a new technology of his day: the printing press. He felt it was nourishing for his monks’ souls (and their work ethic) to hand-copy the Bible.

Meanwhile, ChatGPT is the printing press’s reverse in that it takes all that information disseminated far and wide through the printing press and compresses it back into a well-formed paragraph or essay in answer to a prompt. Does such ease nourish the soul—or numb it?

There are at least three factors to consider when it comes to whether we should use generative AI in church ministry: ethics, optics, and acoustics.

  • Ethics. Am I honest about where I got the content? Am I working diligently, prayerfully, etc.? Do I give credit where credit is due?
  • Optics. What will my leaders and congregation think about my level of reliance on generative AI for content? At what level of AI usage will I be giving others (e.g., students) an excuse to cut corners?
  • Acoustics. Good acoustics in a room are neither muddled nor tinny. That being the case, are my people hearing from me a clear, biblical message—or more of the generic, muddled answer that’s typical of AI? And are they hearing from a human or a robot?

“At what level of AI usage will I be giving others (e.g., students) an excuse to cut corners?”


Of What Should We Be Cautious?

Here are some of the most important cautions we need to pause and consider when it comes to generative AI, especially as church leaders:

The illusion of authority. Unfortunately, AI’s answers can feel like the authoritative answer, even when the chatbot itself admits the risk of misinformation. You’ll even run into completely fabricated “information”—no grounding in truth whatsoever—which apparently and seamlessly fills in gaps within its information.

The biases of the programmer. For ChatGPT’s ethical leanings, see my observations in “What Ethics Do We Learn from Generative AI?” It’s not great news for those of us who believe ethics go far beyond simply whether a particular action brings about good or bad results (i.e., beyond ethical “consequentialism”).

The loss of wisdom. In “ChatGPT Heralds an Intellectual Revolution,” authors Henry Kissinger, Eric Schmidt, and Daniel Huttlenlocher warn, “To the extent that we use our brains less and our machines more, humans may lose some abilities [such as] our own critical thinking.” Marshall McLuhan has described technology as an extension of humans (just as a hammer is an extension of the hand), and as such, the tool can blunt our natural abilities. If AI increasingly does our thinking, what becomes of our ability to reason, which is supposed to be one of our unique traits as humans?


“If AI increasingly does our thinking, what becomes of our ability to reason, which is supposed to be one of our unique traits as humans?”


The irrelevance of humans. It’s worth noting that machines seem discontent with only blue-collar jobs. It’s true that humans will continue to need to frame the use of our machines’ “intelligence” for worthy ends, but will we remember this when it’s increasingly expedient to let machines do more and more of the “thinking” for us?

The allure of convenience. It’s not encouraging to hear teenagers brag that, not only are they using ChatGPT to write their essays, but their teachers have given them the green light to do so.

The blurring of reality. I recently read of an AI machine which consisted of both a “generator” and a “discriminator.” After training on real photos, the machine would generate its own images, which the “discriminator” would then flag as fake. The machine would try again, generating something closer to reality this time, but still not genuine enough for the discriminator. Enough time in this self-contained loop of generation and discrimination, and the machine was able to generate images which for the discriminator were indistinguishable from reality. That’s the level of reality-blurring we are soon up against.


“Humans will continue to need to frame the use of our machines’ ‘intelligence’ for worthy ends.”


The erosion of freedom. Kissinger, Schmidt, and Huttlenlocher continue, “Can we learn, quickly enough, to challenge rather than obey? Or will we in the end be obliged to submit?” Laziness and servitude are that dysfunctional couple that keeps getting back together. The babyish adults in Wall-E seemed uncomfortably close to where we could be headed.

Should We Feel Threatened?

As church leaders, how should we feel about the increasing omnipresence and perceived omniscience of generative AI?

In a world in which a person’s all-knowing research assistant spits out custom answers in seconds, who needs a preacher?

As I asked ChatGPT question after question about religious truth and ethics, I learned a secret. Yes, ChatGPT gives impressive, comprehensive answers about science, history, math, language, technology, etc.

But when it comes to questions about religion and ethics? Most of its answers were not satisfying. When it comes to religious truth and ethics, ChatGPT’s answer basically came down to 1) don’t hurt people, and 2) beyond that, the truth is up to you. Liberating? Perhaps for some. For most, it’s just unhelpful.

So, when it comes to life’s most important questions, most people are just going to get more and more confused. All the while, they are also getting more and more lonely and disconnected. It’s like living alone on an island with an infinite supply of salt water, but with nothing to quench that deepest thirst.


“When it comes to life’s most important questions, most people are just going to get more and more confused.”


That’s why they need the gospel. And you.

Does the emergence of generative AI pose a threat? Perhaps, but it’s even more of an opportunity, because increasingly lonely and confused people are aching for grace and truth.

Get Renew.org Weekly Emails

Want fresh teachings and disciple making content? Sign up to receive a weekly newsletters highlighting our resources and new content to help equip you in your disciple making journey. We’ll also send you emails with other equipping resources from time to time.

You Might Also Like

Ten Common Characteristics of a Healthy Church Regardless of Size or Location

Ten Common Characteristics of a Healthy Church Regardless of Size or Location

Over the past 20 years, I’ve had the privilege of visiting hundreds of churches across the country. Recently, I’ve highlighted several exceptional examples on my website. While there are many more I could have named, these congregations serve as reminders of what it means to be a healthy church. Below are ten characteristics these churches […]

More