Is Artificial Intelligence a Threat or an Amazing Ministry Tool?

Innovative Faith Resources: AI is technology. It feels more human because we chat with it. We ask it questions and it answers. It feels almost sentient. It's not. Hey, this is the Not Alone podcast from the SBC of Virginia, bringing you real-time help and encouragement for your ministry. The goal of the SBC of Virginia is to let you know that no matter what you're doing for the kingdom, you are not alone. My name is Chris Weibel. Well, you can't get on the internet, or for that matter do anything anymore, without an encounter with AI. Artificial Intelligence. If you ask ten people about it, you'll get ten differing opinions. Things like, AI is really helpful, or it's a good way to cheat on your college term paper, or it's going to eliminate all the humans. One of the reasons artificial intelligence is met with so much speculation is because of its lightning-fast emergence. The internet and search engines in Google slowly became a part of our lives, but AI has made a big splash relatively quickly. The thing is, it's probably not going anywhere. So many folks are asking, can AI play a role in ministry? On this Not Alone podcast, Brandon Pickett is talking with Garth Henson, the owner and founder of Aperture Insights in Durham, North Carolina. Aperture is a technical advising firm And one of Garth's goals is to help ministries watch out for problems with and effectively use AI. Here's Brandon. Hey Garth Henson, great to have you with us in the Not Alone Studio. ⁓ it's great to be here, Brandon. I really appreciate the opportunity. And you're coming up from Raleigh, so thanks for being here in person. Yes, sir. And why ⁓ Lynchburg? because you came up today, probably whether you're going to have the podcast or not. Yeah, so ⁓ one thing, my daughter's here at a Liberty First School and she's wrapping up and any excuse I've got to come up and see her, I like to do it. And I've got a few meetings on campus this afternoon as well. So it was nice to be able to see you guys in person and make a day of it. This is great. I'm so glad. And tell us a little bit about your background because it's really varied. It is. It is. And I'll try not to take up too much time. But I feel like I've got to start with, I actually grew up on the mission field. My parents moved to Guam when I was five. And so I grew up in a Christian home. my entire life, kind of being around ministry and active in different ways. And I had an interest in technology, but didn't have a whole lot of exposure. And then the Lord has a way of kind of shaking you up and directing different things. And I was planning to go to school for ⁓ veterinary medicine. I was working at a vet's office in my sophomore year of high school and I developed pretty severe allergies. And so he closed that door pretty quickly and I didn't have a lot to lean back on. And so I actually went to Bob Jones University. I moved back from my senior year of high school, was at school and declared a Bible major. It was the only thing I could think of that was not going to be wasted because I didn't have direction. so at the time, my uncle, who was one of the heads at the IT department at Clemson University, actually kind of mentored me, took me under his wing and said, I think you're wired this way. Let me show you some things. And I started to get into programming and specifically web focused. The Lord kind of just opened an interest and an ability there. And so I was able to get my undergrad in Bible and music. And so I got the art side, I got a little bit of communication training, and then I started into a career actually in technology. And so I actually ended up working back at the university for a few years, got some training, got to audit some programming courses. Through the, I'll try to keep it short, but through my career, I ended up being in Western New York at a company for a while. through many different things that the Lord changed our paths. I ended up out in the Seattle area, worked at Amazon for about four years. And then I joined ⁓ an organization over at Disney and I was in downtown Seattle working for one of the technology groups there. And then reorganization, I followed some leadership over to Microsoft and worked there for a few years before coming back into Disney. And I worked for another five years between Disney and Lucasfilm leading some teams there. To be completely honest, the Lord kind of got me to a point where I don't recall the trigger, but there was something where I was sitting with my wife, and I think it was when I was working on supporting the teams that were bringing Disney Plus and Hulu together on the platform. And something just struck me that if we're here on Earth for a reason, beyond just simply knowing the Lord ourselves, it's to help other people prepare for eternity. And I was sitting, spending every day doing almost the exact opposite. And I was distracting people. My entire career at that point was focused on entertainment and distracting people from reality. And there was a weight that kind of settled in. And it took a few years, but over the next few years, I made a big pivot through, again, no direct choice of my own, but the Lord saw fit to In 2023, I was part of the massive layoffs from the Disney company. And it was the first opportunity to turn to something completely unknown and new. And I pivoted entirely into startup world. And I started working with some small companies, went from enterprise down to just tiny companies. And I started working in advisory role. When AI started up, we started exploring very, very heavily into what does it mean? How does it work? What does it do? Most recently at the beginning of this year, I actually moved into being completely self-employed. And so now I'm actually working alongside some nonprofits and some other companies and helping establish strategy, how we're approaching things, and just helping kind of guide technical strategy in general. So 25 years there and about five minutes. That's a lot, and God has done so much. He has. And I'm sure you've learned a lot because technology has changed so much since you first started in Western New York. I should even go back before that with your uncle. tell me when, because one of the things that ⁓ I'm so glad that you're going to be with us at the Worship Technology Conference at the end of May, May 29th and 30th, you're going to be teaching a couple different things. One is a master class on artificial intelligence and how that can incorporate into ministry. Before I dive into that, let me back up and say, when did you first kind of get into artificial intelligence? When did you first hear about it? What did you think? I'd love to hear this. So that's a great question. And I'm going to have to actually back up about 12 years because it's only been the last few years that we've labeled this technology as AI or artificial intelligence. Several years ago, you would talk with people, you might hear machine learning, computer vision, were some words that were thrown around some technologies at the time. And really all it meant was training an algorithm that would then be able to do human-like detection. So you give it an item and say, identify this. And with pretty high confidence, it could tell you what it was. That was kind of the underpinnings that started the path to what we have now. Now, obviously the... Technology, the underlying technology has changed significantly, but several years ago, my first exposure and where my interest was kind of peaked was while I was working at Disney, we had an opportunity to go on site at Redmond at the Microsoft campus and work with a team that they had that was called the Bot Framework Team. And their whole team was building chat bots. Again, this was 10 years ago, but chat bots that would sit on your website and be able to talk with you in a persona. So. If you wanted somebody who's always cheerful, this bot never says anything critical. If you want specific things. And I recall what was intriguing to me was everything that they would do was boiled down to two things. One was called intent and the other was entities. And really all that means is your intent is what are you wanting to do? What are you wanting to know? And the entities are about what? The items. It's basically your nouns and your verbs, right? And so what ended up happening is we were doing really rudimentary things at the time, like can we have a chat bot that talks as if it's Mickey Mouse, but can answer the question of what parades are at the park today? And so again, very simple. that was, were you in this line of work at the time? With Disney you were trying to come up with. these kind of things? our team was, I worked on the corporate technology team. And so my team actually serviced about 17 different brands within the organization. So we worked on everything from ESPN launches to Club Penguin, if anybody remembers what that was back in the day, it's a kid's website. And so we were working across the organization on exploring some of these newer technologies and finding, they cost effective? Are they reliable enough where we would put them in front of users? And I do believe that because I was working for Disney at the time and the bar of excellence for the company is so high, we were doing a really, really good deep dive into can we rely on these things? Are they going to make a mistake? Are they things we could actually put in front of a customer? And so that was kind of the, in my experience, that was kind of my origination and where things were. And then fast forward several years and I actually made a statement. First time I heard AI being applied to modern technologies was probably three and a half, four years ago. And I saw where things were headed and it bothered me immensely. And I made a statement that I would never work for an AI company. And about eight months later, I found myself working for an AI company. the reason being- Never say never, right? No, no. And I think this kind of getting back to what you were asking is, The reason to it was at a very high level, the approach from a lot of organizations and a lot of individuals is how do I replace people with this technology? Which to me was extremely backwards. My approach is how do I help people spend more time in relational interaction by taking away some of the mundane tasks? And that's where AI really shines is how do we How do we streamline things that I have to, it's just a task. It's part of my daily routine. Can I automate that so I can have conversations with people more and spend time? And so it's not about replacing people. It's about amplifying the side that only people can do. So I don't know that actually answers your question. I want to kind of ask you at the different places, especially like at Lucas, Microsoft and Disney, what were your focus there? when you were at each of those places? So when I first started at Disney, I was actually leading the login and registration platform that was across all the brands. so if you logged into ESPN.com, Star Wars.com, Club Penguin, again, back in the day, those were all the same platform. And I was leading that group, that stack, if you will. So that was originally. And then I moved into a corporate architecture team, where our job was to start exploring kind what you're describing. What's up and coming? What do we do with it? What can help us for cost effective help us in other ways? And there was a, I would say a lot of companies say they're on the quote and you'll see this in things that the bleeding edge, right? Like, ⁓ stuff comes out, we're using it right away. And we were taking a measured approach of what could we be using within the next 12 to 24 months. And so we were exploring that edge, but with the, you know, downstream perspective of What's going to be affordable? If anybody's explored AI, you realize it's not cheap. It's quite expensive to use in bulk, in volume. And so it's things like that. Like when does it become cost effective? How do we start exploring those? When I went to Microsoft, I was working with a team where our responsibility was actually ⁓ governing international data centers. So we were managing all the onboarding and leasing of new data centers. So. several billion dollars worth of revenue going through these data centers. And so our job there was not AI focused at all, but it was how do we navigate a new data center in Germany through 50 pages of questions, including can you keep the floor temperature to a certain degree because we can't have the heat too high. And so I worked in that group for a bit. was there, I want to say I was there two and a half, three years, and then moved back into Disney. when I was asked to come back, it was on the streaming side and Disney Plus had just finished the larger portion acquisition of Hulu. And they were starting the project to bring the platforms together. And so I actually was leading a small team, only about eight or nine of us together. And our job was actually to provide tools and technology for the 350 web engineers that were working on bringing these platforms together. And so that's what I was doing. And then again, to your point, fast forward just a little bit further. We got that team pretty well established, took a couple years. And then I was asked to come across and actually hire some people to amplify a team at Lucasfilm. I'll give you just a little enterprise breakdown. Lucasfilm sits alongside an organization called ILM, Industrial Light and Magic. Most people have heard of that. You've also heard of Skywalker Sound. Just so happens my son loves those like a little. documentary on that. that was a good one. I have to say I liked it too. No, that's great. the three organizations, Lucasfilm is properly a subsidiary of Walt Disney Company now. ILM and Skywalker Sound were intentionally split out as individual entities because they service organizations outside the company. They work with Sony and they work with Paramount and they work with other, they provide post-processing and editing. There's a small team, about 75 people, that sits on the intersection of those three companies, between Lucasfilm, Industrial Light and Magic, and Skywalker Sound. It's called the Advanced Development Group. And that was the team I was asked to join and start working in an area of essentially, without getting too complex, my job was to come up with a way to share data between the organizations and Disney. in a way where each organization could still maintain control of what was theirs and only share what was necessary, specifically around the realm of we were working to get interactive content into the set-top Disney Plus app. you could, instead of just sitting back and watching, the idea was you could actually play a game right on your TV and interact with it. And so that was the group that I was with there for a little while. It sounds like you were trying to be a pioneer, like your group is pioneering new things, looking on the horizon. Tell me if I'm wrong here. You're looking on the horizon and you're saying, okay, is this going to work? Let's, let's dive in. Let's test this before we roll it out to the masses. We've got to find out all of the bugs in here or if it even works. you're always out there scouring your team to see what can work for these three. And then obviously for beyond that. Is that right? When I was at Disney, we actually formed a new team called the Emerging Technologies Team. And that was our charter, was to look 12 to 24 months down the road and look at what is happening today that we want to advise or influence the organization to adopt within the next 12 to 24 months. So that's exactly it. So just like a preacher would say, well, that is all my setup for the message. All of this is the introduction. Well, that was the introduction for what we're about ready to talk to because you have been out there. You have been looking, you've been scouring on behalf of all these companies, but now we're talking about ministries. As I talk to pastors and church leaders and church staff, and whether it's at a worship technology conference or whether it's just they call me up and they say, Brandon, help us understand. then I have to turn to guys like you and say, hold on, Let me get back with you. I've got to call Garth. So Garth, help us understand what, and this is a big question, maybe we can just narrow it down a little bit. What is out there, not 24 months, but what's out there now that churches can say, you know what, we can't live in fear of this. We serve the creator of the universe. All good things and perfect gifts comes down from the father of lights. We can use these gifts to help spread the gospel, to disciple. So Garth, help us understand what's out there now that you feel in your experience. And that's a big question. is, it is. But I mean, I know we have all these things like ChatGP, and Claude, and Gemini, and all these things. Is that really what we can use, or is there something that we're not even thinking of? ⁓ man. How long do we have? You have five minutes now. I love this question. ⁓ This is actually one of the major things we're going to talk about in the breakout session as well. ⁓ good. I like a tease. Just sign up right now. Right. I'm to have to do my best not to give too much away. But I think one of the things we've got to do, and you touched on it both in your question earlier and in what you just said, is technology changes so rapidly. I used to tell people that if you're getting a computer science degree in a university, by the time you're out, what you've been taught is about two years behind. ⁓ man. Not necessarily a problem, but you've got to be aware of that. Right. AI, in the last three years, has taken that, what I would have said, a two-year cycle, two-year window of turnover in technology and without exaggerating, I think we've seen between three and six months that speed of iteration now. And so what would have been potentially eight years, eight to 10 years maybe of technical growth and platform changes, we've seen in two years. And so we've seen four or five entire cycles because of this technology. I think where we can find ourselves very quickly is fear. Because of that, it's like, well, I can't even keep track what's happening. For so many decades, things had stayed pretty static. And then I'm to go back to my childhood. It seemed like the 70s, 80s, something around there, technology started changing. And it's speeding up and speeding up. I'm going to really age myself. So I remember when my grandmother and my granddad got a microwave and that was like, whoa, but they got it. But I don't even know they really wanted to use it, but they got it. And then all of sudden then it become part now we can't imagine life without it. So here we are. There are, think now from what I'm kind of getting the lay of the land, there are still people who are fearful of AI, but more and more want to start trying and start seeing how we can use it in ministry in a responsible way. Theologically, ⁓ not giving up your calling to turn that over to AI, but how can we use it and what's out there that pastors can use? ⁓ great. Not to divert the question whatsoever, but I do feel like we need to take one step back and we need to acknowledge AI is technology. Because of the way it's been presented, it feels more human. in nature because we chat with it. We ask it questions and it answers. It feels almost sentient. It's not. And I think one of the biggest things and a reality check we have to just establish is it's just a technology. How it is used can be evil or good. And you mentioned aging yourself a little bit. I'm old enough where I remember when we were first having battles with churches about do we even use the internet? Do we put a website out there? And I remember fighting a battle for probably my first five or six years of my career of sitting with church leaders and saying the demographic, at the time I was saying between 25 and 35, the majority of them, the first place they go when they make a move or do something is to the internet and they do a search. I was saying, they go to Yahoo or something like that. And we had an entire battle of this is a technology. It is not inherently evil. It is used for lots of evil, but the technology can be used for good. And so, I think that is one of the kind of pivotal moments we're having with ministry now, is it's like, okay, we recognize and we see some of the bad that's coming out of it, we see some of the things that are happening. How do we responsibly, how do we ⁓ accountably use the things we've been given in a way that we are still taking responsibility and ownership for the things we have been given to steward. We can talk pretty quickly and come up with these different guidelines, but these are hard. These are not easy things. We could even see the line in the sand and it's like, well, which side of the line does this particular topic fall on? And so one of the things that we'll be doing in more depth at the event is to to break down, okay, well, let's understand as a technology, you mentioned several, chat, GPT, Claude, things like this. And the one thing that I'll make sure to break out again, I'll touch on it now, we'll do more later, is those are actually just tools that have been built on top of the AI. So we think about, like I talk to people all the time and they use the term synonymously. Yes. Right? They're like, ⁓ have you tried chat GPT? And I'm like, yes. Like, ⁓ like Claude? I'm like, no, those aren't the same thing. And people will use chat GPT as this is AI. One of the things we'll do is take a step back and say, well, let's understand there are two parts of AI as it's been presented. One, you'll hear people talk about the large language models. That's the brain. That's where the information resides. What sits in front of it is that chat interface. It's how are you interacting with the brain? And so we kind of break that apart a little bit. Yeah. So when you went back and you said really probably should be called computer learning rather than artificial intelligence. It was. Like the original algorithms and things, it was machine learning. it continues to and continues to. But tell me this, because I was talking with somebody recently, actually my brother-in-law, who was reminding me. that we think that it's going out into every single website in the whole world and bringing back all of the data and it's not. He was saying it's actually quite narrow. Is that true, depending on the model? So this is where we split hairs a little bit again. There is the interface layer. So if you're interacting with ChatGPT or Claude or you're on their website and you're doing things, that's different than the model itself. The model sits underneath it. The model... is static, it's been trained, it only has access to the information it's been trained on, and the model itself can do no self-learning. ⁓ What happens is the interfaces in front of it have access to the internet. So if you say go explore such and such and evaluate this content from this website, if you do that through ChatGPT or Clawd, any of the anthropic models there, it will do what you're asking it to do. but it's because the interface will go scrape the content and then in addition to the question you asked, it will take all the content of the website you pointed it to and it will package it all together in one prompt or one message down to the model. Now the model has all of the information available to do some evaluation on, but it can only evaluate based on its training data set. And so again, it is a little bit fuzzy. If you don't understand the distinction between the model can only make evaluations, I'm trying to choose my words carefully because it's not deterministic, but it can only make evaluations based on its training set. This is why you have new versions regularly coming out. Anthropic Clawed just released Opus 47, I believe we're on now, just in the last couple of weeks. Well, only a month and a half ago, 46 came out. And this is because they're continuously training the models on more and more data. And depending on what it's been trained on, it can do certain things better than other things. And so again, not to get too technical, but when people come to me and say, which of the providers should I use for this particular task? My answer is, well, yes. Because if you want to evaluate something and you just wanted to, let's say I have a document that it's 150 pages long. and I want a two page summary, personally, I use two different models in combination for that. I'll go to ChatGPT and use their lightweight analysis model, is for many, if anybody cares. And that does an extremely good job at distilling information and summarizing. It does a very, very poor job of creating human speech. And so I'll take the results from that. Then I'll go over to one of the Claude models because those seem to excel at human communication better. Now by itself, it wouldn't be able to analyze my document as well. So I'll take the analysis from one, but then I'll go into the other and say, now I want you to summarize this document or rewrite this document in a form I can digest more easily. And by using those together, I actually get the best result for my use case. So again, there isn't a one size fits all. The way I try to explain to people is a model, a large language model, has been trained on as much information as they could find to put into it, yes. But it can only answer as accurately as the information in the prompt has narrowed the data. I'll give you one of the examples we'll use in the breakout session. But you'll hear people talk about AI is very, very good at Prediction. Now this isn't weird like prophecy. What they mean by prediction is what's the next thing coming. And what I mean by that is if I go into a room and I say, okay, everybody wants upon a blank time, almost everybody's gonna say time. But if I went into the room and I said, you know what, there is a company that's an organic foods company started by Jennifer Garner called Once Upon a Blank. Well, those who know are gonna say, ⁓ farm. Well, once upon a farm is actually the correct answer I was looking for. Now, my question to the room is, which answer is correct? It both depends on the question. The context is what determines what's right or wrong. And so by the model answering time, it's not wrong. But I didn't give it enough context to get a correct answer for my use. If I give it organic food company, this information, called once upon a something, it's going to come back with, farm, because you told me what you're looking for. It's very similarly to if we think about it at a university. At a given university, if you think about the university as the brain, that is the model. You've got access to all kinds of information, but it's not until you find the coursework to train you and the specialization you're after and you finally graduate that your presence as that specialist becomes valuable. And so it's actually taking all of this access to information and trimming it down to a very narrow lane and training you in that. Now you have extreme value within your lane. But if you've been trained to be a neurosurgeon, I'm not gonna ask you to change out my circuit board at home for my electrical panel. I would not even trust you to do that. Does that mean you're not intelligent? Not at all. But your specialization is in one field. And so, what I'm discovering and what I'm encouraging a lot of organizations and in this case, ministries to do is don't think about how do I use this massive brain? Think about, are there narrow specializations that we can tune it to be able to do very small, narrow things extremely well? The answer is yes. One of the things that the workshop will actually be looking at together is one thing I've done for a couple ministries is if you have a public email inbox. This is probably even more so for nonprofits than for churches directly, but if you get several hundred emails a day, I guarantee you that scattered through that email inbox, are some very high priority critical emails. It could be somebody struggling with suicide. It could be any, and those you don't wanna miss, but you've gotta read every single one of those emails and you've gotta crawl through the ones that are, sure, maybe it's a, great message on Sunday, which is good, that's encouraging, but I would argue the criticality of responding to that is lower than somebody who's legitimately having life struggles. It's not that either one is less important, the priority, the criticality is different. And so what we've done is we've built a filter, if you will, where every message that comes into that public folder gets run through a model. And the instructions to the model are, you are not making a judgment call. What you're doing is, here are my 10 classifications. I just want you to tag emails. with the classifications that apply to the content in this email. Why? Because then I go to my inbox and I sort based on the classifications, highest criticality, lowest criticality, and now I can at least have a little more confidence that I'm addressing the right things first. Again, does not reduce the complexity or the importance of things, but it's a way to give me visibility to make my decision faster and more accurately. And think that's the key is that how do we leverage the tools not to take decision making away ever. And that's my hard line in the sand is I am not trusting AI to make decisions on my behalf, but it's actually a lot better at reading 150 page document than I am. It remembers more. And so to distill that and let me read a three page summary, I can make decisions faster and more accurately. than I could if I had to ingest that entire document myself. So again, lot of, I know I just kind of brain dumped for a little bit, but there's a lot, a lot to consider there. And you are actually now, you're creating, what is that that you're creating for ministries? I don't know exactly what you would like. For me, it's just, it's an AI application because it's not, a GPT also has the mentality that it's an interaction point, it's a chat interface. we do some of those too. We'll do specific GPTs that are trained on doctrinal statements and theological boundaries and things like that. And again, this is one of those touchy areas. if you have the right boundaries set, it can actually do some pretty incredible things to help start fielding, hey, did I say anything in this outline that is outside of the theological boundaries I've already set for myself? because what is it good at? It's good at finding patterns. And so it will actually be able to help review sermon notes and things like that. For example, a pastor may want to use an illustration from somebody, but they don't know the background. And I've heard this where they'll pull a quote from somebody that says, I love this quote. And they pull it in. And then they didn't realize that this person has theological, they don't have the theological boundaries. And so they're using somebody that they wouldn't espouse. And so they wouldn't have known that, there's a way to actually check, does this person have this? One step before that even that I've actually had to talk to a few friends of mine about is they've gone to some of these chat interfaces to generate sermon illustrations. And the problem I have here is now we have completely made up things that sound great and they bring an emotional, you know, response, but they're not true. you One of the ways we proclaim the good news of Jesus Christ is investing in the next generation. That's why NextGen exists through the SBC of Virginia. Through dynamic resources, events, and opportunities, your youth group is not alone. From leadership training to discipleship, you have the resources at your fingertips to lead the next generation well. Learn more by visiting our website, sbcv.org slash NextGen. So a pastor will go, hey, look, here's my sermon outline or here's my topic. And I need some illustrations about this topic. And it brings you back, but they're not true. Right. Oh my. And so that's one of the things is that, and again, another thing we'll cover a little bit more in the session is that the AI, it is a computer and it's going to do what you ask it to do. If I tell it to give me a story about Sarah, it's gonna make up a story about Sarah and it's gonna be very confident about its story. I don't know, you could come up with different questions and you can even ask about, or you could instruct it to do something and if it doesn't have accurate information in its data set, it's still gonna generate content. You'll hear the term hallucination. That's what this is. When people are using AI and they talk about, it's hallucinating. That's exactly what they're talking about is it reaches the bounds of what it actually has access to. So it can't fulfill your request, but it must fulfill your request, so it's going to start making up information to return to you. with some of the newer models, because the contexts are massive, you can have literally, you can put the entire Bible as a document into some of the contexts for these now and have conversation, but I'll go find such and such in this particular, you know, within these books and it'll have, it's amazing how much you can do with them. But, they're very confidently wrong sometimes. And you have to, I'll give this quick example, my dad's been working on a couple books, ⁓ he's been, he wrote one ⁓ on the Advent, it's a devotional book, and then now he's working through the book of John. And he's been using it to do historical research on what was the alignment of the planets and the stars in that time period based on the location in Jerusalem or Bethlehem or What were the weather patterns like? And what's interesting is it's actually really good at being able to go back and look at all the data we have and say to the best of our knowledge, it would have been like this. And again, he's not going to say it was rainy on this day, but you can say, well, during that time period, most likely in that area, this time of year would have been temperate. Yeah, rainy season, dry season. Right, right, with some confidence. But if he were to say, tell me what happened on the third Thursday of the month of Nissan for this particular region, it's gonna come back and try and come answer that for you. Humanly, hopefully, we're gonna say, I don't know. AI is not. AI is going to confidently answer your, you said, something for me. And it says, OK, I'm going to do this for you, whether I have it or not. I'm going to, that's And some of them will start to tell you you're reaching the limits of my knowledge, and it will try and do some type of conceding. Here's an example of that. My wife, Wendy, and I, we've been looking at different ancestry. And we said, well, hey, can you help us And we went into Chat Cheapie T and we said, who was the mother and father of such and such? And it says, there are, can't accurately or definitively tell you this because of the records, but from what I can see, here are options. Well, that was pretty good because it didn't say, say we, it know, it knows for sure. I wish it would do that with everything. It's getting better because I think definitely getting better. One of the early examples of probably doing things before we should have is when Gemini was first released. I think it was the very first, maybe the second version. There are screenshots that are floating around where people were doing Google searches and there was a ⁓ screenshot I've gotten in one of the talks that I did where somebody had asked, what should I do about depression? And in the AI response, it said, some users recommend jumping off the Golden Gate Bridge. And what was happening is it had access to distill information and it was taking the best matches of the actual statement and returning the context that had gathered. It wasn't claiming do this, but it was the first response was some users. It was Reddit responses. a very agnostic response to. Right. But it was encouraging suicide for depression. And again, Those that was very quickly pulled back and they redid something. So again, there are checks that are starting to become more careful, but it's still, again, it's a technology. It is not sentient. It has no experience. It has no emotion. Throughout scripture, we're commanded to not just get knowledge, but to get it with wisdom. That to me is the biggest gap between AI and humanity. AI is smarter than I will ever be. It has access to more raw information than I will ever be privy to. However, it has no wisdom. It has no experience. It has no soul. One of the biggest challenges I would give ministry leaders is discern the thing you're looking at. Does it require wisdom or is it a fact check? is simply a knowledge thing. Example, I've been leading worship for 20 years with my wife at different churches as we moved around the country, and I would have loved to have AI years ago when I was trying to schedule, you know, multiple different groups and our choir schedule and our concerts coming up and all these different things. I can now put in, here is my entire list of every song I've got access to. I can add stuff as we go. Here's my entire schedule for the last three years. Here's historically what we've done. ⁓ and by the way, here is my preference. I want one hymn at least. I want maybe one or two modern hymns or songs. Maybe I want a worship chorus kicked in there too. But here's my preference of service shape. ⁓ and by the way, our church, maybe we're leaning reformed. And so this is the structure of my service. I can give it all that information now and say, and by the way, don't repeat a song more than once every four months. And then I can say, generate the next 12 months of a proposed schedule for me. That's just factual. That's knowledge-based. And it does a really good job at doing things like that. Now again, it doesn't then know what is the correlation of the rest of the service and what's the sermon gonna be. So there's still a human factor before you want to take an output like that and go and minister with it. But to have 80 % of that busy work, if you will, of the scheduling and the all I'm doing is looking at my calendar and when was the last time we did this? ⁓ I can't remember. Let me go look it up. To have at your fingertips the ability to say, here's all of my rules. What should next week's service look like? I talked to a number of worship pastors recently about this topic. And they were back and forth a little bit on that because a couple of the worship pastors were like, well, I feel like that, I want to be led by the Spirit in, because I don't know, just because we haven't done a song in four weeks or maybe we did it last week and the Lord wants me to do that again. So again, what you went back to, which was the facts and the decisions, don't let AI make all your decisions for you, for sure. It's not a decision. especially when you're talking about ministry, especially when you're talking about the Spirit-led service. But it can tell you, which I really like, give me the frequency of these songs. Give me how many songs that we've done, you know, like this or this style or whatever. I love that. Also, I like putting my ⁓ praise team and band or whatever, and then put in, hey, build me out a schedule based on this this person's going to be out for this month. This person's good. It's almost like an HR kind of thing. administrative side, right? Yeah. I need that. Right. Because if Johnny says, Hey, I'm going to be out for these days, put that in there. And then I'm bringing on new people, but I don't want to tax them too much. You know, that kind of stuff. Absolutely. love that idea. Yeah. That's where what you're touching on goes back to when you said, what do you build? What do you do for ministries or for organizations? is a lot of people don't think about that as AI, because what do you have? Well, you have a form and you send it to all of your worship team members and you say, can you just fill out what your vacation is gonna look like for the next six months? That's the kind of thing we've seen in the web since late 90s. It's no different. What we will do is exactly what you just described is behind the scenes we'll say, okay, I don't wanna have to do the footwork of doing all of these calculations on the fly. Can we just have access to one of these large language models that knows how to do scheduling well and say, okay, here's the schedule, just make it line up. And so taking a layer under the hood, that's one of the other things where I am back and forth. And I've so far, I've settled on the line of I am not yet comfortable in a ministry setting, having in business, we'd say customer facing, but visitor facing. AI interaction. I personally, this is a conviction thing that I'm not going to tell anybody they're doing wrong. I personally cannot in good conscience put a chat bot on my church website, an AI driven chat bot, because I've seen them fail too many times. And in that one instance where it does fail, is it worth the relational connection? with a visitor, a member, whoever it is, is it worth it to say, cool, we got a chat bot? I have for myself kind of built this six pillars. It's a philosophy of how I approach technology in general. And one of the things we'll talk through is to try to help addressing the fear first, right? Because we've kind of done the whole gamut in this conversation already is, You've got fearful, I don't want to do it. And then you've got the other extreme, which is, ⁓ this is so cool, put it everywhere. And what's interesting is if you study through scripture and you actually look at it, there are examples of times where the Lord actually blesses those who use technology well. You have things like Noah and the Ark, where he's commanded to build a structure that was unlike anything seen to that day. But I learned something as I was preparing for this conference that I had not actually seen before is that If I mentioned King Uriah, most people associate him with Daniel in the year that King Uriah died. It's like, okay, but did you know that during his reign, we have a little bit in Chronicles about, very little, but we have some about his reign. But one of the interesting verses, I think, is one verse that talks about essentially siege engines that he had built with skilled craftsmen that were sitting on the corners of the walls in Jerusalem that essentially could hurl boulders. I read that. Wait a minute, this was 350 BC. This is Jerusalem we're talking about. And this is like cutting edge technology that in the same breath of he was blessed of the Lord, it's talking about the skilled craftsmen he brought together to use technology to defend the city. And again, it was just one of those moments where I went, the Lord is not anti-technology. He's given us the ability to do it, but we've also been given very clear responsibility. and accountability that we have to balance. I think that's where the hardest part of this conversation comes. It's a lot easier to answer in a business world. Yes. Because the mistakes are quantifiable with money. We're talking about dealing with people's lives and their souls. That means for me personally, the line is way back here somewhere of what I'm willing to do. And so I give that a little bit of context because I... A lot of people are like, ⁓ you're just AI. I'm like, no, no, no. I'm actually very, very cautious of what I'll do, but I've reached a point where I don't believe it's any more a fearful thing or should we it's a no. How do we use it? So I'm going to hit pause right here. We're not stopping this conversation. We're to hit pause. And the reason being is because Garth is going to be with us for two days in May, May 29th and 30th at Liberty university in Lynchburg, Virginia. He is going to have a master class starting at one o'clock on that Friday. And it's called from prompts to power tools, AI for ministry leaders. You have got to be there. And the way to do that is go to worshiptechnology.net and go on and register right now. And then he's going to have a breakout or two on Saturday. He's going to be with us. It's going to be an amazing time. By the way, those master classes, I don't know if you saw them Garth, but they, we've got incredible master classes that day from some of the country's best. It's amazing. was definitely a little, little daunted to look at those. I'm going to be alongside of so. So just to, we've got, if you're looking for project management, Bobby Puffenberg is going to be there building trust with your team and congregation. Cliff Lambert, free your natural voice, Andy Christman, planning to presence. We've got LJ Mitchell there. And so you can go through all of these. Charles Billingsley is going to be talking about the balanced worship leader. pro presenter masterclass. I mean, I could take that one right now. And then also a deep dive into church audio with Drew Hester. So we've got eight of these masterclasses just on Friday. We've got a worship leader and even, I want to say technology leaders come on too, to this dinner, bring your spouse that Friday night and then all day on Saturday. Well, Garth, as we wrap up here, I know we've, gone through a lot of different things over the last 50 minutes or so. Just scratch the surface. Just scratch the surface. But what can you leave us with? It's just some encouragement as we look ahead here. I'll reiterate a couple of the things that we've said, because to me, those are the core things are, number one, acknowledging that the Lord has given us an opportunity to live during this time. We have a ability to reach people in a way we have never had before in history. The footprint that we can have in people's lives, and I think where I would leave the biggest challenge and encouragement is that can only ever be a personal thing. Because if you think about our job is to bring them so that they can have a personal relationship with Jesus. And technology, no matter how far it goes, can never replace we're the only Jesus some people are gonna see. That can absolutely never be replaced with technology. And I think that would be my biggest encouragement is let's find ways to be able to start setting aside some of the, maybe tactical is the wrong word, but some of the routine tasks so that we can actually focus on those relationships and really, really try to draw people closer to the Lord. With the advent of any new tool or technology, there are always doubts and concerns. And in a ministry context, we have to be extra careful to make sure the use of new technology is handled in such a way that it always is God-honoring. One small slip-up can have big consequences. So, what about your ministry? Has AI been an interest? Or is it something you're steering completely clear of? Email Brandon Pickett and he can help get you connected with Garth Henson. Brandon's email is bpickett at sbcv.org. And if you want to really dig deep, plan to come to the SBCV's Worship and Technology Conference this May 29th and 30th at Liberty University in Lynchburg, You'll get to meet 30 worship and technology professionals for breakout sessions and master classes on the most effective ways to use technology in your ministry environment. Garth Henson is presenting a masterclass on AI for ministry leaders. Again, the Worship and Technology Conference is May 29th and 30th in Lynchburg. Go to worshiptechnology.net to sign up. The Not Alone Podcast is a production of Innovative Faith Resources. If your ministry is considering new digital media options, well, you're definitely not alone. Innovative Faith Resources will walk you through the tough technological challenges that come your way. To find out more, visit sbcv.org. That's sbcv.org. These resources will disciple, encourage, strengthen, and challenge you and your church family to live daily for Jesus Christ.