Tech & Kids: Understanding AI Podcast Episode

For each of us is being able to understand where it’s drawing its information from to understand where to apply Ai. We’re just going to use that label, where to apply AI in our lives. And then also some areas where we need to be careful how we apply it. Very careful. And then there’s probably some areas where it shouldn’t be involved at all. 

Welcome everyone to Breaking Bread, the podcast brought to you by Apostolic Christian Counseling and Family Services. Excellent as always, to have you along. Matt Kaufmann is my name, and I’m really pleased to have a guest here, David Virkler. Welcome David. Great to have you here. Thanks Matt. It’s good to be here. 

Well, let me give you a little bit of introduction to David and then it will just make a whole lot of sense of why I’ve asked him to come in and address and talk about the topic that we want to talk about and that is technology and in particular, AI. So, I know David because he runs our technology security here at ACCFS. 

And so, we rest easily at night because David is taking care of things on the cyber front. That has also allowed David to interact with us as an agency for training and how we can use technology wisely and safely in a very wild west cyber reality. I really appreciate David, his ability to explain things, his ability to walk through and make hard things easy. And that’s what I need in this tech space, David. So, welcome great to have you with us. Say a little bit about yourself, a little bit about your background, and elaborate on anything there that I’ve said. 

Sure. Yeah. So grew up here in the area, Central Illinois, and through the years God led me to different places. And so, my wife and I now reside in the Akron, Ohio area. That’s where we attend. I have five children there. Yeah, I’m just so thankful for the opportunities that I’ve had. I have been in technology my whole career. 

And it has been really interesting the last five or 10 years, especially as it just continues to get broader and broader. And so, one of the things that I will just say up front is I would not tell you that I’m an expert in AI or really any technology at this point. It’s just so broad and so deep. 

And so, I am looking forward to sharing together and exploring this topic. Yeah. The playing field just changes all the time. And I have a tremendous amount of respect for those of you in the tech world making a living on technology and the pressure that you must be under to stay on top of a thing that is always moving. 

Yeah, it is hard to be an expert as we try to navigate raising kids in this technological world. I want to put my finger on some heartburn, some difficulty when it comes to helping my kids think through their work skills or interaction with a boss or about money and budgeting, I feel like I’ve got a thing to say because I’ve done it before. 

Right. And a lot of parenting is bringing up children because, hey. I’ve done this before. When it comes to technology, I haven’t been through it any further than they have. Right. Which I think puts us a little bit on our heels and often kids are ahead of us and we’re trailing behind. And I think it’s a unique space in which to parent. 

And so, what I am looking for, David, is to talk about artificial intelligence. Okay. Because it seems to be exploding and is really impactful. But I also want you to help educate us. I’m going to be listening for some information, but I also am interested in precautions that you would see as it concerns AI. 

How about ways that we should be engaging others, ourselves and our children on this topic. And then finally redeeming some really exciting things that are coming out of the field. So, those are some of the things I’m listening for. I want our listeners to be able to hear, because I think this can provide a great deal of value on those matters. 

Does that make sense? Sounds good. So, let’s start with this informational piece. Help us get underneath the hood. What is AI simply and what is AI more deeply? Sounds great. Yeah, so it’s kind of a buzzword right now. We use AI a lot in our conversations and most accurately, we’re probably all using it wrong, but we’ll drill down here. 

So artificial intelligence has been around for decades and at the simplest definition, it would just mean intelligent machines. Right. And so, this has been something that’s been in progress for decades. And so, if we start with a big circle of artificial intelligence, then we’re going to drill down further into that. 

Inside of that, there’s some pattern recognition technology, which many have heard of called machine learning. We drill down further into that, and there is what’s called neural networks. So, this is a subset of machine learning. Neural networks or deep learning. And we’re going to talk a little bit more about this. And then further inside deep learning and inside these neural networks is something called large language models. LLMs most commonly known to people today as chat GPT. So that would be a large language model.  

Well, I think that’s fascinating just to see these at different levels and to recognize that smart devices is another way to say intelligent devices, right? Yeah. We know these can be everywhere from the thermostat in our house and in many other places. So, this idea that we are using AI has already integrated into our lives, whether we know it or not. Yes. And one of the interesting things about this for us parents is that we could say that it’s the next big thing since the internet in the technological world. Would you agree with that? I would agree with that. Yeah. As far as its impact on the percentage of the world. Right. AI would probably be, and again there I said AI. 

Really, it’s the large language models because that was really the breakthrough that happened with Chat GPT a couple years ago. I think it was November of 2022. Yes. If I got that right, there’s a large language model called chat GPT that became available to the public. And that’s really where we saw this AI buzzword take off and we started seeing all these services built around AI come to the market with advertising. 

So really, when we say AI, it’s actually that very specific. So, let’s linger there a little bit with language models. So, what’s happening there? Yeah. How does that work? Yep. So, with a large language model, what’s really going on when we’re searching in Chat GPT is, it is doing its best when it’s looking to generate an answer, to predict the next word. 

So, to generate the next word. So, these models, these neural networks, well, that even happens when I’m texting, right? When I text, it’s like do you want to say thank you? Yep. Very much. You know what I mean? It’s kind of new, so it’s similar to that except it has way more context. Yeah. And way more information to pull from. 

Yeah. Right. So, these neural networks that I talked about have all been pre-trained based on a body of data. And this is really important. This will come up later, I’m sure, but this pre-trained data has come from all of the publicly available information on the internet. And so, I think it’s really important for parents not to view AI as magic or a black box. 

And so, ironically, you talked about this as something a lot of parents haven’t been through before. Yeah. Like maybe we have with buying a house or some of those other things, but ironically, one of the things that I think parents should be doing is to figure out how to teach this to your kids. 

So, understand at least at a very high level that you take away the mystery of it. I perceive that you’re saying there’s a real shadow here when this becomes separate from humanity or something like that. Is that correct? Right. And so, that’s why one of the areas of danger that we’ll probably get into a little bit is when we start to look at it as a being or we start to look at it as having some knowledge that nobody on Earth has. 

Right. That’s really not what this is about at all. And so, I think one piece of guidance for each of us is being able to understand where it’s drawing its information from to understand where to apply AI. I’m just going to use that label, where to apply AI in our lives. And then also some areas where we need to be careful how we apply it. 

Very careful. And then there are probably some areas where it shouldn’t be involved at all. Yeah. Right. Okay. Well let’s take those one at a time. Where does it draw its information and are all AI models pulling from the same data set, or is that separate and different? So, Chat GPT is a pre-trained model, so it’s pulling its information from books that have been electronically transcribed that are available on the internet and user generated content. 

So, think social media, think forums, Reddit, things that are publicly available that have had humans generating the words that are on those pages, you know, like medical journals, right? All of this information that’s been accumulated over the years. It’s somewhat ironic if you recall over the past five years where we’ve come to know that when there’s a free service, we are the product, right? And so, it’s our information that they were using to help generate revenue for their organization to keep functioning. So, one of the red flags that comes up in my mind is accuracy. Is that still a lens that we would be thinking of? 

And if the data pool has got errors in it, it runs a chance then that the AI is going to produce error. Yes, very much so. So, when we look at how it’s trying to generate, going back to that concept of how it works, it’s trying to generate the next word to give an answer that it thinks you want, right? 

So, when we look at how it’s doing that, it’s looking at the inputs we give it, and then it’s considering weights within its neural network and a few other things. But when it’s trying to generate that next word, it’s important to know that it is not necessarily using truth as one of its requirements for output but rather using probability at some level, right? 

Yes, absolutely. Yep. Using probability, what’s the most likely word that would make the most sense next? Right? And so, that’s where we run into these things called hallucinations, which when AI will generate text, right? Generates an answer. Really, it’s just helpful to understand it’s really just generating one word at a time to build out that answer, and it’s generating text. It’s not a full thought. It’s generating text. And so, just because it says it doesn’t make it true because truth was not a requirement in its capabilities of generating the text answer, it’s going to have a persona of a great deal of consciousness or meaning because of the way it’s read and the way it comes off. 

Yes, exactly. It’s very fascinating. I just scratched the surface and the little bit of research that I did, but it’s very fascinating how much work they went into breaking down how language works because you think about it. I put a question into Chat GPT and how does it know what I’m asking? Right? 

How does it know what I mean? Like what’s the thought that I’m going for? Yeah. And so, it has to break apart and figure out which words are important in that question, which ones aren’t, and which ones to run through this neural network with which priority for it to actually come out the other side with something that was useful to me. 

Interesting. Am I right to say that you can have AI generators that are constrained and the data that they’re pulling from? So, for example, a business says, here’s our AI model, and it’s not pulling from Reddit, it’s not interested in Reddit. And so not all AI forums are equal. 

Is that true? Yep, that’s true. So, you can have AI engines, which are really those neural networks that I was talking about. You can have engines that are trained on different data sets. And so that’s important to know when you think about the data that you’re putting into this, right? You want to know whether that’s being used to further train this engine or whether it’s contained to just the instance that you’re working with. 

And so, this has become very important in the medical industry, right? Where potentially searching up symptoms or something like that in an AI engine, those are going to be self-contained. And so, there’s a caution there for our kids too, is to understand that what you put into an AI engine, Chap GPT or others, we need to be aware that potentially that’s actually recorded and then used further on. 

And, you know, one of the things about AI is that this particular talk could potentially be out of date within a year. Right. And so, there’s a lot that we don’t know. And so, I think in our conversation here, we’re going to look for some, some timeless principles too. Yeah. Right. Because we don’t really know what the future’s going to bring in this world of AI. For us to understand how to use this well is to understand where the data is being pulled from. So, if I’m asking a medical question and I’m in Chat GTP, that’s pulling from the whole web. That’s different than the AI model that Mayo’s might be using. Yeah. Right, right. Because we do hear of really bad advice, for example, or AI putting out some very harmful things to people searching for answers. Yes. Yeah. So, considering the source of data and understanding how it works, actually does two things for us and our kids. 

One is it helps us use AI better for those good things that we want to use AI for. But two, as I talked about earlier, it takes away the mystery and helps us discern on those other areas that are maybe middle ground or a no-go zone for using AI. And it helps us discern in those areas when we really understand how it’s working. 

Again, at a high level, it doesn’t need to be at a graduate degree level, but just understand where the data set came from, how it arrives at its answer, that it’s not always correct, that it’s really not going for truth. That is something that they’re really focusing on for these next versions of Chat GPT. 

So GPT Five is I think the latest as of this date and trying to get more truth is one of the priorities for them. Okay. So, knowing where the data is coming from, knowing that if I contribute to this, is it, or is it not going into the engine, are some things you’ve mentioned that we should know. And then you said, knowing how to use it, having some places that you would have some caution about using it and some places off the table. I’d love for you to step through some of those areas as well. How do we use it?  

Well. Yeah, using AI, there are a few things that it’s actually really good at in its current state. Again, this may change in the future. One of the things that it’s really good at is writing. It does a great job at piecing together words helping with tone. It’s excellent at collecting large data sets and being able to analyze patterns. One of the things that machine learning does well is pattern recognition. So, going back decades, those are some of the things it was starting to get good at back then. And so, looking for patterns and data will be a thing for our kids. They will need to know how to use IOL for many future jobs. As a teacher, I have been confronted to really rethinking what I want out of students because of AI. 

If it’s just the right answer, it can’t be anymore. Because there’s no mystery in the right answer, right? And so, I find myself really up against rethinking what my role as an educator is. What am I trying to educate and what am I trying to. achieve in the minds of students? Does that make sense? 

Yeah. And I think a lot of us are having to rethink those things because of the service that Chat GPT, for example, does. Yeah. Education is highly impacted by AI as we’re calling it, because the teaching is not about presenting the information anymore, right. The information is at their fingertips. It becomes application or real world. Right. Chat GPT can’t tell you how to apply the knowledge when you’re working with a certain type of person who has a family history that you only know about because you interact with them in real life. 

Right. I’m curious if there are some cautions that have that come to mind saying, you know what, we might be out in front of our skis in this area or that area as it concerns AI. Does that make sense? Yeah, there are probably the two areas that come to mind for cautions, one, especially for our kids, would be the emotional arena. So, one of the things that’s taking place with AI is several services out there that would be your friend, and it’s an AI powered companion. Yeah. And it’s just a very dangerous space. So, one of the things we need to consider is that this is the response from AI, from these large language models. 

It’s trying to give you the answer that it thinks you want. And so, when we think of our kids going to chat GPT and looking for, you know, maybe for example, why does nobody like me? Or something like that. That’s very deep. Emotionally, it becomes the echo chamber of all echo chambers. 

Well, because it is looking to give you the information that you are looking for. And so, I think of our role as parents as much as we want to and should be friends with our children, that’s not the only role that we have. There are times when we need to say words that don’t feel good. Right. And so, if our children are starting to use this as a coping mechanism or a place to find friendship, a place to find somebody that understands, it’s going to feel good to them. Yes. Because this is a friend that will never hurt your feelings in a day and age where isolation is rampant. 

Yeah. And youth anxiety is at a height where we may be uniquely ripe to fall into some of these ditches. Yeah. The word that was used was sycophantic, which is used to describe behavior that is self-seeking to gain personal advantage. And so, this can be used to describe a person as well. 

But this actually started way back with social media many years ago and the concept it’s designed to mimic. Again, what the person is looking for and their intent in that is that it’s more useful, but there’s also a piece to it that self-serving keeps you coming back. Yes. Okay. Okay, so now you’re really getting to the economy of it. Right? How are things running? So, for example, if there’s a dollar behind it, if there’s an economy of producing a product, I’m sure everybody who’s in manufacturing can tell you that a bad model is to produce something somebody doesn’t want. Right. That just doesn’t sell anything.  

So, you want to produce something that people want. And that people come back for. And so, if having a keen eye on what you want and how I can better deliver is what it’s running on, you can see that it can be a recipe for concern when sometimes what’s best for us is not what we want. 

Right. And so, that ties into the second area of concern which would be our spiritual wellbeing. Right. And so, teaching our children to properly use these tools is fascinating. We go back to those of us that perhaps grew up in a search engine age where we learn to Google everything. Our children are growing up in an age where everything’s going to be fed into AI. 

And so, there’s a piece of this where in our spiritual walk, it really is this worshiping of intelligence that we’re seeing right, where we’re living in a society where almost all knowledge of mankind is available at your fingertips. And so, considering what God has to say about some of that and how that relates to our faith, how it relates to curiosity in a person. You know, I think they’ll be speaking about our kids as potential future jobs. I feel that there’s going to be a real need for people that can actually have original thoughts and have creativity, and to be uninfluenced by maybe what the data says. 

Yeah, that’s interesting. Now you’re speaking to some timeless qualities of creativity and critical thinking that need to be matured and grown alongside this model. Is it possible that if we don’t, this model could undercut those things and produce somebody who doesn’t critically think or produce somebody who’s less creative. 

Only time will tell. I guess I’m just curious in your opinion on that. Yeah, I think it’s something we need to be mindful of. I think of the example at our house, you know, you’re sitting around the dinner table, and a question comes up, does your hand shoot in your pocket for your phone to find the answer? 

Yeah. Right. Or do we just let ourselves talk together and share and wonder and think and work it out in time. Yeah. Instead of immediacy. Oh, yes. That’s another big piece of this. Yeah. As we all know, there’s so much to our walk of faith that we don’t get instant answers for. And this is a habit-forming tool that gives us instant answers. 

And so, we may have our boundaries in good places but in our professional life, we’re used to getting all of our answers instantly. Yes. But then over here in our spiritual life, how do those habits start to train us and train our mind. And even how we feel about not getting an instant answer. We need to be alert to what this is producing in ourselves. Yes. And this is probably something that has been, we could say simmering, for a long time, right? As we think about the past 20 years of adoption and technology really being integrated into our lives, more so it’s crafting our habits without us necessarily choosing that. 

Right? And so, it’s an opportunity for us to really step back and say, where should the boundaries be? And are there areas of my life that this is either starting to control or we’re just unsafe? And so, I think this is an opportunity for us as parents, not only to exercise, but also to teach our children of those whom we want as mentors in our lives. 

Who do we want to speak into our life? Right. And certainly, as believers, we want God through the Holy Spirit to be speaking into our life, but AI is somewhat like going to some world conference where you have an expert on every topic there. And would you just send your kids to the building and tell them to ask whatever question they want? 

Yeah. Right. We probably wouldn’t be okay with some of the answers that they would be getting. And they also don’t have the tool sets for discernment. Knowledge is at our fingertips and yet wisdom is another matter altogether. And I’m not sure AI has a lot to offer for wisdom yet. But the mentor you talked about in the Spirit of God certainly does. 

And often wisdom comes through experience, time, slow path, and to have wisdom come alongside knowledge is going to be the best way forward, right? Absolutely. Yeah. So, on the positive side of this, when we can use AI in a safe and productive way to amplify our God-given talents, it can be a beautiful thing that brings glory. You live and breathe technology. That’s your daily life. And I’m curious, what insights do you have about God himself? How has AI illuminated aspects of God or made him grander than maybe what you previously thought of? In what ways has God been contemporary in our AI world and not archaic.  

Because I think if we don’t carefully steward this, we can get the idea that God is archaic. Yeah. There’s one of the Bible accounts that sticks out to me when it comes to ai, looking at it from a little bit of a Tower of Babel moment where you had God’s creation and all these people saying, we’ve achieved it, right? Aha. We’re going to be self-sufficient, better than God. We’ll reach the heavens. And so, there are real challenges with AI where it’s going. There may not even be enough power in the world to run some of these things. 

And so, just to keep it all in perspective, you know that this is all information living on physical computers that are located on an earth that God created. All this stuff is being computed with processors that are all made with elements that God created. Right. Just keeping it all in perspective that it would not take much at all to shut all this down. It really doesn’t. And you know, it might even look different than we could ever imagine, too. You think of the tower of Babel. Yeah. You know, who would’ve thought of shutting that down just by changing language. Yeah, right. Didn’t tear down the tower. Yeah. Just changed the language and it brought everything to a halt. 

Yeah. That’s really fascinating. It is flimsy in that way. Speak to humanity because I think sometimes that is a little bit at the crux. It seems like there’s a blurring of humanity and Christ has got a lot at stake in what human is. And so anyway, I’m curious if that generates any reactions. 

Yeah, there are certainly a lot of questions that this brings up that we just don’t know the answers to. But I think one of the things that it brings out is it reinforces and highlights the benefits of our human relationships. And then as we go search the Scriptures, we find that was God’s intention. Right. Accountability is there. Sharing the gospel message and even Jesus coming to earth. Yeah. Right. There was a reason. Right. There was a reason, and it has to do with that human-to-human connection. I think it’s a beautiful opportunity to talk about the soul. 

Because that is something that sets apart mankind from machine, right? So these human looking robots, if time continues on, will likely start to be in our world but they have no soul. And so, it just highlights and gives us opportunity to think about and reflect back to the humanity point. 

Yeah. It reflects our short time here on earth. We can think about when this body dies, just like a machine will fall apart and eventually get to the end of its useful life, that there is a part of us that is undying and that sets us apart and that is different. Oh, I love that. Being a good student of what it is to be human is going to be important to the next generation. 

You wonderfully brought Christ into the conversation of AI. We can see him as part of the conversation and superior to the product itself. And if we don’t steward this, our kids can separate that God is not in this space at all. Yeah. And that’s not going to bode well, right? 

Yeah. AI has no soul, is not able to have a conversion experience. Right. All these things we think about with our soul and that’s why I feel like these are conversations that we wouldn’t have had to have 20 years ago. Yeah. But now we have the opportunity, and I love that. I think some of the Scripture is going to really explode with new meaning with the backdrop of AI in our tech world. And I think that’s exciting to think about. Let’s go to this redeeming part, and you’ve already tipped your hat, David, at the usefulness of AI. How are you using it in your everyday life? 

Yeah, so largely using it professionally, using it for work, and we are very careful about the information that we feed into it, but I’ll confess that I find myself still using search engines over AI. And it’s just an interesting generational thing because there are certainly kids in college who will absolutely go to Chat GPT over Google. 

I’m admittedly fairly conservative in my use of AI, but it’s largely habitual. I don’t have the muscle memory to do that, to go there. And then something else too is that we’re going to continue to run into it more and more without trying. So, we go to the doctor’s office, and they ask if it’s okay if this AI agent listens in on our conversation so that it can take my notes for me and can summarize what we talked about? We have that in our business world as well, having these meeting note takers or AI bots that are listening to the conversation for the benefit of compiling information and generating notes and summaries and stuff like that. 

So, I think we’ll just continue to find more and more benefits, and I think something key for us as parents is understanding that this will be our children’s default. It’s what they’re being taught in school. It’s what they’ll be expected to know when they enter the workplace. And so, the opportunity really is ours to safely explore it for the intent that we can help them navigate it.  

I really like that. Anything you’re excited about as you think about AI doing some really cool things? There are a lot of things. One of the things I’m actually really excited about is the education space. I think it’ll provide a lot of opportunities for everyone. I think it’ll also allow people to digest information in the format that works best for them. And so that you don’t have the dichotomy of the way the teacher presents the information not matching the way the student wants or needs to receive the information. You don’t have that fight, if you will.  

So. students will be able to digest information, and then you could say that the teacher’s role becomes more important, but also more valuable when it comes to helping them apply that knowledge to real world scenarios, relationships, implementation, which technology historically has done that, freed you up for more. Yeah. And so, I think that is a challenge for all of us educators and otherwise to think, okay, what more can I do now that AI is providing that one service. But I think there’s more. I was talking to a brother who’s using AI with language and sharing the gospel in some tremendous ways. 

I mean, you think about that Great Commission to take the gospel to the whole world. AI’s doing tremendous things in that space. And so, there’s a lot of good that we’re going to see come out of it. Yeah. Real time translation is just amazing. Being able to have those conversations with somebody in a different language. 

Even in our brotherhood, if there’s somebody who struggles with a second language, how beautiful it is to be able to communicate more freely. Yeah. Right. And to allow the heart to flow. Yeah, that’s excellent, David. I really appreciate it. Thanks for coming in. I think all of us parents and grandparents, and even for myself, because I need to be educated and that’s what happened here today. Educated but also with a tremendous lens of hope and purpose for engaging our kids in this matter.  

We appreciate the work that you do, David. Thank you. And to all of our cybersecurity folks out there on the front lines, it’s so important, so needed. I appreciate the work that you do for us. It’s a pleasure to serve. Thanks, each one, for listening in. God bless you. 

We’ll see you next time.