53. AI in L&D: Overreliance, Resistance, or Exuberance? with Timothy Frost
60m 41s
In this podcast episode, host Jess Omley and guest Tim Frost explore the current state of AI in L&D, framing it as a "tri-linked crossroads" of three conflicting attitudes: exuberance, resistance, and overreliance. Exuberance is driven by vendors and executives who see AI as a quick fix for cost savings, while resistance arises from fear of job loss and the effort of learning new tools. Overreliance occurs when professionals let AI replace their own critical thinking, such as using AI notes as personal thoughts. Jess illustrates AI's practical potential by describing how she uses ChatGPT for personalized dietary advice, highlighting how AI enables instant, iterative, and non-judgmental learning—a model she believes should inform workplace L&D. Tim emphasizes that AI is a tool, not a replacement, pointing out that even advanced AI still fails at basic physical tasks. He argues that L&D must use AI to advocate for its strategic value rather than be seen as replaceable. The episode challenges listeners to reflect on their own attitudes toward AI and how these impact their work, encouraging a balanced approach that leverages AI's efficiency without losing human judgment. The goal is for L&D to evolve from reactive request-takers to proactive strategic partners.
[MUSIC] Welcome to the L&D Must Change Podcast, a conversation dedicated to uncovering the changes that learning and development professionals must make to truly solve our modern talent challenges. My name is Jess Omley. I've been searching throughout my 25 plus year people development career for the ways we can improve our profession, both big and small. I've discovered there's no single person who can shift how we work. It must be a collective effort. So let's learn from each other and raise the bar and our profession together. This is L&D Must Change. Hello everyone. Welcome to another episode of the L&D Must Change Podcast. I'm your host, Learning and Talent Development Transformation Leader, Jess Omley. Before we get into today's episode, let me ask you, is your learning or talent development team the best kept secret in your organization? And really, you can see the talent challenges, you can see the performance gaps, the turnover, the struggling managers, and you know training isn't the answer. But most of your work is still reactive. You're stuck taking requests instead of driving business forward by improving performance and solving talent challenges. Meanwhile, the most effective learning and talent development teams, their strategic partners, they're in the room for the big business conversations. They're tying their work directly to performance and outcomes and the business can measure their impact. This gap is not a capability gap. This is a clarity and language gap. And that is exactly what my L&D strategic business partner team assessment and development roadmap solves. It shows you where your team stands across nine key competencies and it gives you a practical plan to close the gaps inside of your real work. So you can walk into that C-suite with confidence, with the data and the language you need to stop being a secret and start making a real impact. If you want to learn more, reach out. Let's talk. Okay, I want to start today's episode with a couple of questions for you to think about. First one is this. How would you classify your feelings or your attitude? You're not a touchy, feely person. Think about your attitude towards AI today in this moment right now as you listen to these words. How would you describe your feelings or your attitude about AI? And the second question. How do those feelings or how does that attitude impact your work as an L&D professional? Does it impact how you show up, what your day looks like, how you problem solve, how you do your work? Ask myself those questions and personally, I think I'm at a place that's mostly curiosity paired with a teensy bit of skepticism. My curiosity is around what can AI do? How can I use it in my day to day work? How can I continue to use it in the best way possible? My skepticism, that has to do with really the way we're throwing around AI, like it's a magical answer to so many things. Like every software vendor now says they use AI. What does that mean? Or companies are cutting jobs because AI can do it. Really? Are we there yet? I don't know. That's my skepticism. But then I thought about it a little bit more and I realized the way I use and interact with AI now as opposed to a year ago is drastically different. How about you? Using AI now for me is part of my everyday life and I have seen the tools get better over the past year. Then I wonder maybe part of the state of AI and L&D is defined how we the humans think about it, feel about it, our attitudes about it. And maybe we are at a point where all of us are doing that differently or thinking about it differently. Our attitudes are a little bit different from one person to the next. Maybe your answer to those first questions I asked, how would you classify your feelings and attitude about AI today at this moment and how does that impact your work? Maybe your answers are drastically different than mine and I think that's okay. Recently my friend Tim Frost and I, we were chatting about the current state of AI in L&D and how that impacts our work. And he made this really interesting statement. He said AI adoption in L&D is at a weird tri-linked crossroads of exuberance versus resistance versus over reliance. Tri-linked crossroads like picture a triangle with three different points and the points are exuberance, resistance and over reliance. AI and L&D we're in the middle. That statement may be stop and think. Think about all those things and those questions I just asked you. How are people in L&D classifying right now their thoughts, their feelings, their attitudes towards AI and how is that impacting our work? So naturally I had to have Tim come out in the podcast and discuss this in more depth because AI is one of those things that is changing our lives. Whether we want it to or not it's here it's happening and it's impacting the work that we do. So let me throw another question at you one more. What was the last thing you used AI to help you do? The last thing was it the right tool for the job? Where did you as the human step in, where did the AI do the work, was it effective? As I record this the last thing I did with my AI was ask it what I should eat for breakfast. Seriously it's not as innocent as you think. See I'm trying to be healthier. Balance my diet a bit more. Balance the fats, the carbs, the proteins, the fiber, all the things you hear about but I'm not an expert so I don't really know how to balance all those things. I know what they are but how much should I eat? What constitutes those things? Now it's not something that's entirely new to me because overall I do eat pretty healthy. I also have a few dietary restrictions which can make things complicated and so I've worked with dietitians and nutritionists in the past and they've been great. They've provided me with guidelines, things to try, even full recipe books but I've been feeling sluggish lately and I didn't want to fork over cash for a nutrition coach. Plus I didn't even know where to start. I didn't want to try to find someone, read reviews, figure out who to talk to, take a chance on whether or not they're going to be able to help me. It all seemed like it would take too long so I started asking chat GPT what I should eat. I put in my goals, I put in my restrictions, I told it when I had energy in the day, when it dipped, basically I gave it all the data and then I asked it for some breakfast and lunch ideas. It came up with several. I tried one, I was still hungry afterwards so I went back to it and I said hey I was still hungry afterwards what should I do instead and it gave me suggestions and I've been continuing this conversation every breakfast and lunch for the most part meal that I have for the past couple of weeks. And now I have something that I never had when I paid nutritionists and dietitians to help me in the past. I have access to an instant non-judgmental dietary assistant one who knows my goals, my restrictions and will answer endless questions at any moment in time without judgment or a hefty bill. None of the nutritionists or dietitians I've worked within the past have been available at any moment like that for these types of questions that I've had to save them up. I've had to ask them when I have an appointment, maybe I can email them and then I have to wait for a response. So in the meantime I've ended up guessing or just going back to my old habits. That's how I've been using L&D or using AI. How does that relate to L&D? That was the question I wanted to get to. I am literally learning in the moment in a personalized way by asking certain questions when I need to immediately apply the answers and then iterating as I go. If you ask me, this is the new norm. I remember a decade or so ago my kids were teenagers and I've always paid attention to how people are learning outside of the formal structured classroom. Because I've always wondered if we learn differently outside of work, why are we trying to do something old school within our work? So a decade or so ago my kids were turning to YouTube whenever they had questions. And I was. I started turning to YouTube and Google search and it made me wonder why were we trying to fit learning into a formal box at work when we choose alternative methods to learn in our personal lives. So I started leading strategies that mimicked personal life learning more than what started to feel like old school formalized learning in our workplace. The way we preferred to learn had started to evolve from formalized
to informal in the moment videos. And now I believe we are evolving again. And I think in L&D we should pay attention to that. What are the implications of AI use in our personal lives where we choose what, when, how to learn? What about you? Where are you at with AI? And how is it impacting your L&D work or how should it be impacting your L&D work? Part of what was intriguing to me when Tim said we are at a weird, tri-linked crossroads of over reliance versus resistance versus exuberance, was that I think I experienced each of those when it comes to AI even within the same day. I might be over reliant on it at one moment, maybe my food plan. I might be exuberant about what it might be able to do at another, and I might be a bit resistant, resistant to believing all the hype in another moment. But also curious and a teensy bit skeptical, as I mentioned earlier. I have a feeling you might be too. So here's the challenge for you this week. Think about your feelings or your attitude towards AI at any given moment, really, really. Like how do you feel about it? How is it impacting your work? Where might a good place be to try using it more or differently based on what you're experiencing and what you're seeing around you? Take a listen to this conversation between Tim and I as we unpack this a bit more and then let us know your thoughts or tell someone else, ask them the same questions, keep the conversation going so we can keep learning and growing together. I'm hoping that this episode that we're really seeing now it's gonna be kind of outdated in the next year. That hoping that all of us in learning and talent development have grown leaps and bounds in our use of AI and those around us have done the same. All right, with that challenge, let me formally introduce Tim Frost to you. Tim Frost is a digital learning and performance consultant who partners with organizations to solve real business problems, not just build training. With a background in education and enterprise healthcare, he specializes in transforming complex systems into intuitive, high impact learning experiences. His work sits at the intersection of strategy, technology, and design. He helps organizations modernize their learning ecosystems, improve adoption, and create experiences people actually want to engage with. I hope this conversation sparks as many thoughts for you as it did for me and I can't wait to hear what those are. So without further ado, here's my conversation with Tim Frost. Hello Tim, I'm so thrilled to have you here today on the LND Must Change podcast. - Well, thank you for having me, Jess. It's great to be here. - Yeah, I can't wait for this conversation. So before we jump into all of the goodness that we're gonna talk about, tell us, who is Tim Frost in this world of LND? - Well, that's a great question. And when Tim Frost figures it out, I will let you know, no, I'm kidding. But I am Tim Frost and I hail from Orlando, Florida. And I am about a five year practitioner in LND in the healthcare space. And I am a former high school educator, like so many of us in the industry are. I like to tell people I did a 10 year sentence in the public school system. But I think sometimes that experience makes us the best educators, best instructional designers, best school professors. - Absolutely. I don't think there's any bad experiences if we learn from them and move and grow and move on, I guess. - 100%. - Yeah. Okay, so I always kick off our conversation with the same question, which is, how do you believe that LND must shift or change? That is such a great and dynamic question. And I think for me, LND needs to change with organizational appreciation and adoption. And I know that's a loaded question or a loaded statement rather. But I think that's all too often LND gets flattened or in the case of our discussion today, is seen as replaceable by tools like AI or things. And I think that LND needs to do better at using the tools to advocate and lift themselves into the space to add that value that we all know as practitioners, as instructional designers, as learning leaders, that we offer organizations so that when those leaders are looking at, where can I save a book? Where can I save sense here and there that LND isn't the first name that comes to mind? - Mm-hmm, yeah, very much so. And you and I, when we chatted earlier, let's just dive into that tools piece. Like let's just go there right away, Tim, because you know that's the thing that so many people are talking about. And I feel like we can't actually have a podcast anymore if we don't say AI, 'cause then we've like missed out on checking the bingo card or something. But you and I had a very specific conversation and you said something that I thought was so fascinating. You said AI adoption in learning and development, it's at this weird tri-linked crossroads. Exuberance versus resistance versus overreliance. And we have to find the happy comedian. So before we get into more of that, what's exuberance, resistance, and overreliance? Can you give us an example of what each of those look like? You know, I think that exuberance is, whenever you're at a conference or an expo floor, you know, I think in the last two to three years, everything's gone from, here's my learning tool and here's what it can do for your organization too. Hey, my learning tool now has AI. And you know, I think you walk down, you know, the I-L-A-D-E expo floor and it's like, you know, there's not gonna be a vendor that doesn't have something that's AI related. And you know, with the exuberance of organizations, you know, you have executive leaders that see the, you know, the value of ROI or whether it's head counts or time or productivity or whatnot. You know, so they're constantly looking at, you know, how can AI help us save dollars, you know? So I think there's a lot of exuberance. And I've seen that in my own work across the industry of where can I save dollars or like I was saying, you know, everything has AI and in it, you know, I think the resistance comes in from the human fear factor of, you know, like, not really knowing what the intention of the technology is or what a leader's intention of technology is, you know, and so I think naturally, you know, that resistance of, well, if I don't know how to use AI, then, you know, I'm good, you know, like my work is still gonna happen, you know, I don't need to, you know, train a tool to replace me, you know, for lack of a better term. And I think the overreliance comes into play where, you know, we don't go to meetings or we rely on the AI notes to fill us in, or we start using those AI notes as our own thoughts, you know, so that's where I've seen you know, people are struggling with the AI overreliance. And, you know, I'm just gonna put it in chat GBT, or I'm gonna put it in co-pilot, and it's gonna formulate my thoughts and that's gonna be my thought on it. Instead of actually using our educated brains to formulate our own thoughts. - Yeah, I can see that. - Sure. - So all that put together is like that weird triangulation of like, we're caught in the middle of this, of like, there's the, is AI, is chat GBT agents gonna be the next instructional designer at my organization, you know, is my executive, what's my executive's really intention with, you know, this AI adoption model, and then also like, well, AI can just do it easier for me so I can just do other things, you know, and that's the weird triangulation of where we're at in the middle. - Yeah, it is. It's almost like these, these different perspectives that are swirling around this, like, sometimes we're really excited about it, other times we're thinking, "Oh, I don't know about this," and other times we're like, "Oh, this is awesome. "Let me just ask chat GBT because it'll give me the answers "and then I don't have to think about it." And you're right, that all of those things swirling around together do make for sort of a weird point in time, potentially. So let's dive into each of those just a little bit more. Let's talk resistance first. So you said we've, well, let's just say we, maybe you and I, maybe not you and I, or our L&D colleagues, we're a little bit resistance towards AI. What does that look like when it shows up? How do we know somebody's a little bit resistant to AI based on what they say in a meeting room or when I'm having a conversation with them? I mean, I think, and I've seen several instructional designers across the industry really shy away from the good parts of AI, and really it kind of stems from that fear. It's like resistance and fear kind of together, but they shy away from wanting to use AI because they see it either as lazy, the lazy man's way out. Again, it goes into that fear factor of, is this going to replace me? Do I need to know this? Am I training it to do my job? Am I going to be out on the street? And then also there's the age old technical, something new and technical comes into play. And I might be later in my career, or I don't have time or energy to learn something new. And so I'm just going to keep doing what I'm going to keep doing.
And I think that's where that resistance is primarily coming from. It's kind of its own weird side triangulation of this weird Ross roads of fear and not wanting to learn and adapt. But in the same breath, I've also seen instructional designers really look at AI and save BOOK who amount of time trying to do all the good things of AI, whether it's image generation or storyboarding or scripting, things that used to take us hours or days or weeks to do that can now be done in seconds. But that almost can pair with the resistance. Because if I'm seeing that what I used to do can now be done in seconds, that depending on how I view that and what my perspective is, that could be the AI can do this in seconds and it takes me four or five hours or a couple days, what does that mean? So you mentioned there are those L&D professionals who are worried. AI's coming from me. What's the response you give someone who says, AI's coming from me. I'm freaked out about it. Okay. Have you ever seen those videos on like social media where the robots trying across the street? And it hits the curb and falls down. That's my response. AI is a tool. And I think that the people, the executive leaders that are out there in the world screaming, we're going to replace headcount with AI agents or AI bots or we're going to reduce headcount by 50%. I think that they are the ones that are the naive in the world. Because if a CEO that's making XD million dollars a year is out there saying, "Well, AI can replace 5 million dollars worth of the learning team." Okay. Why can't AI replace a six million dollar CEO? It goes both ways there. I say all that to say that as far as the reality of the situation is AI is a tool. It is not the be all end all. Yes, AI is getting better every single day. It is advancing every single day. The things that you can do with it now that you couldn't do with it six months ago is mind-blowing. At the end of the day, we still can't get AI robots to walk straight and not fall over. There are a long ways away from even I think remotely allowing AI to take over these critical roles such as learning and development. Because there's always got to be the human factor. Even now in any AI course that I ever produce or teach or work with professionals on, the underlying statement is always validate your output. Because even now, even though AI is not hallucinating as much as it used to, it's learning more. You still got to validate it. Otherwise, we're going to get into this, talking about weird triangulations. The AI is teaching the AI while the AI is grading the AI. The AI is going to live in this whole little world and the humans are just going to be here looking at each other. I think the key to that is as L&D professionals and getting back to that resistance statement or resistance thought is we shouldn't be learning as L&D professionals how to use AI. In the sense, we should be learning, and I'm going to steal this quote from, believe it, his name is Jeremy Utley. It was up the Adobe Learning Summit that he said it. But he said, we shouldn't be learning how to use AI. We should be learning how to work with AI. How can you use that AI tool to support your output or support your work? You're not working or excuse me, AI is not, you're not using AI. You're working collaboratively with it. So it's not a replacement. It's an enhancement. Exactly. Exactly. I think I hear you saying, when we're talking about being afraid that it might replace us, hold on, it's not going to completely replace us. But there's a huge opportunity here for us to enhance the work we do and do it even better, working with the AI. 100%. I'll give you an example there. I like to tell people that I use chat GPT sometimes as my therapist. I'll be in the car and I'll have this, we all have our best thinking in random places. My mind is when I'm driving to work in the car. I'll plug the chat GPT voice on the car, and I'll just brain dump with her. I named her Sandra. Sandra is my chat GPT agent. So I'll just brain dump with my chat GPT agent Sandra. She just levels me out. She tells me, "Okay, you want to get to this point with this project? Go ABBC and D." I walk into my day having thought about whatever I need to think about or brain dump on, whatever I need to think about. Like you said, it's an enhancement of what we're doing, where I have one of the smartest friends and co-pilots, if you will, on the side, that can level me out, that can help me organize my day, organize my thoughts, and make me be more present in the work that I need to do for that day. 100%. When you say AI is an enhancement, I support that 10 fold wholeheartedly. It is an enhancement to what we do every day. So in your case, it's making you better because it's priming you. You're like coming in with the right head space from the beginning because you've already taken that time to maybe reflect on, ask questions, prepare for your day, even before you hit your desk and like you have 50 emails to answer and all the other stuff starts moving. I can see how that's hugely helpful. 100%. That's even getting out of the realm of how you can use AI to triage those 50 emails. That's another fun little trick. For those that are resistant to AI that might be listening. These are the moments when you just dabble in and try to deal with it. If you come in and you get flooded with emails and you have co-pilot in the Microsoft environment, for example, ask it to triage your emails. What do I need to respond to first? Not to give Microsoft a co-pilot a plug, but they even have a feature where you can schedule a prompt in the morning. So you could schedule a prompt at 7am that says, "Summarize all the emails that came in from me overnight and tell me which five I need to respond to the first 10 minutes of the office." And it runs that prompt at 7am, sends you an email in your inbox. It's the first thing you see in your inbox. I look at that and I'm like, bam, bam, bam, bam. That's where I need to go. That's fantastic. And I realize you're not plugging co-pilot, but the reality is a lot of the workplaces co-pilot is the one that's built into the systems that we're already using. And here's one interesting thing. There's a lot of admin stuff in AI that bogs us down. What you just told me was one of the things that I hope happens in our profession, where we're able to help ourselves not get so bogged down with all that admin stuff that happens every day, but move through it a little bit more quickly, prioritize accordingly. And now we can get to the bigger projects. Because I mean, I'm guilty of it, Tim. I find myself, okay, I'm going to just answer a few emails and then two hours have passed. And I realize, I haven't even gotten into my big projects for the day. So I love that's just a super practical tip for anyone to even just try it out and see what happens. 100%. So all right, we talk about resistance. But let's like swing the pendulum all the way to the other side. Maybe there's three sides. So another side, which is the overreliance. Other corner corner. Okay, the other corner. So you told me earlier, you think overreliance on AI can actually hurt us in our learning and development profession. It can hurt our learning outcomes. It can reduce the quality of our products. It could negatively impact our team's capabilities. What do you mean by that? Maybe explain that a little bit more. Where does that? One of my greatest fears as a learning leader, especially when we got articulate AI was, are we going to become a churn and burn shop? And I know I realize there's a lot of organizations out there, a lot of learning professionals out there that that's the role. You're a churn and burn. You get the request. You do the needs analysis. You produce the course. You send it on its way. And perhaps that's why articulate AI and like the document to course feature that they have, which I look at as a real asset, by the way. Perhaps that's why it exists. But I think my biggest fear with tools like that and with the overreliance of AI is that kind of what I said a little bit earlier, are we going to become the type of leaders or the type of learning professionals where that's all we're doing is we're taking the documents that the SMEs give us and we're going to plug it into the AI tool and then output a course and publish it and be done with it. And so really you have all this AI generated stuff. And yes, it's getting better. It's getting more engaging. But that's where I look back at that human factor that we were talking about. We're the college educated professionals. We are the ones, we're the humans that are presumably producing courses and content for humans.
to consume. And so, yes, the AI tool helps us to output the content quickly and efficiently more than ever it has before. But the danger of that and the overreliance of that is if we get bogged down by pressure, by timeline, by project needs or leadership push or just volume in general, do we get down to that point where it's just AI produced course after AI produced course. And then the worst implication of that is the human factor of what does that do to our learners, what does that do to our businesses, what does that do to our organizations. And again, I want to acknowledge that the content itself is getting better, but it's never going to replace things that we can do as instructional designers. And I'll give you another example. I have yet to see an AI produced course, fully AI produced course that has been gamified in a meaningful and quality way. Not to say that you can't use things like vibe coding or what not to produce games for a course. There's a time and place for that. But if you have a content bucket, if you will, we are the ones that need to take that. Yeah, you can use the AI tool to synthesize it and build it. But let's use that to still not get away from those core instructional design concepts, such as gamification that really help build engagement with our learners. So we still need to have our learning expertise, critical thinking hats on. That's what I'm hearing you say. We can't just say, okay, the AI can do it. I'm super busy. I'll plug this information from this meat. That probably also came from AI, by the way. I'll plug this AI information into the AI and then create the course for me. And then I don't ever look at it again. I still need that critical oversight. Exactly. And the interactions. I tell my team all the time, I'm like, I am the, and this might be controversial. So this might be like the Jess headline. Tim Frost says, multiple choices crap. I hate multiple choice questions. And I know there's a time and place for them. And I know there's time and place for a knowledge check. But I would much rather have a more meaningful interaction to encourage our learners to consume the content versus a multiple choice knowledge check. And I think that maybe it's not as controversial as I think. I don't know, but I'm glad you're not. I mean, I definitely agree with you. And I do not like a multiple choice either because it just brings us back to so many things from school and whatnot. But also as a learner, I don't like multiple choice. And I sometimes, this is probably another topic for another day, Tim. But sometimes I think, why do we create stuff that we wouldn't want to do ourselves? So that to me is one of those things. Like, why would I get, I would never create a multiple choice test for myself. So why do I want to create it for my learners? If there's a more meaningful way, that will help them to embed the concepts more into their memory. So is just simple recall. Yeah. Exactly. And I think, I think kind of getting back on track for sure. I like how you're bringing us back on track. Thanks. Thanks for hosting the podcast, Tim. That's good. That's good. That's good. That's good. Listen, listen. Where were we? No, anyway. But coming back, you know, like, I think that, you know, the tools like the file to file the course feature is so fantastic in the sense of like, it's going to give you that framework. But then, you know, to your point, like as humans, as the critical thinkers, as the designers, as the innovators, we go back and we figure out, okay, the AI put in, you know, our famed multiple choice knowledge check. Let's replace that with, you know, maybe it's a vibe coded game knowledge check. Maybe it's a simulation type thing. Maybe whatever your flavor is. But we go back and kind of make those edits in there throughout. Can you just for those who might not know? Because I think vibe coding is starting to bubble up in our lexicon. But I don't know that everyone really understands what it is. Can you give us a brief, a brief overview of what that means and how you use it? Yeah. 100%. And I think that, you know, that's speaking of podcasts topics. That's a whole nother, you know, yeah, probably is changing. That's a whole nother series. But edits, edits, core of vibe coding is using AI to develop either applications or tools that you traditionally would need to use either JavaScript or HTML coding or some other type of code based language to develop those just within either an AI tool. And you might use that AI tool to write the code. So for example, like you might use AI to write a JavaScript snippet that you would plug into storyline for an interaction. I've done that before with confetti interactions. And I think storyline has in their AI assistant, they actually have the JavaScript now enabled, but it'll do that. But or, you know, you're using tools like lovable where you can kind of do like concept to app, you know, where you use plain language. I want to develop, I want you to develop a web page app that converts PDF to word docs, you know, and it goes through the whole background and writes all the code to do what you need it to do. So using AI tools to write code. Yeah. Thank you. Thank you for explaining that. There may be listeners who are sitting here saying, yeah, I've already been using that. But there may be plenty who've said, I've just heard this word and I don't really get what it is because I'm not that far along. So appreciate that level setting. Okay, we talked about resistance, we talked about over reliance, but we have not talked about exuberance. Sounds very exciting. So it does sound very exciting. But I think when you and I chatted about this earlier, we talked about exuberance, particularly around like senior leaders and execs. And I suspect because having worked with many execs and been in many senior leader circles myself, it's a push for efficiency and cost savings probably at the top. You know, because we're in these like economic times where everything's really scrutinized, we're trying to figure out how do we do more with less, how do we work leaner? And so it's really tempting, I think, for a leader, senior leader who's not in the day to day to say, all right, this is a viable method for making that happen. We can be leaner, we can do more with less it's not personal, but we need to keep our business profitable. But it can seem like that's disconnected from reality, right? So I'm thinking about this. And you know, I'm thinking about the the exec who says, just use AI for that like, or the senior leader who says, or any leader who's just like, you know, what we can just use co-pilot. Let's do that. So how do we push back against that without becoming the LND team that's anti AI? You know, the best place I've kind of seen this over exuberance. And I think there's a couple different kind of looks at it. You know, I'm building off of what you said as far as over exuberance in the like the savings or the consolidation line. I think that you really have to think about, you know, number one, what is the AI tool capable of? You know, not all AI tools are alike. And I once had a colleague tell me, you know, the same thing you just said, well, co-pilot can do that. And then, you know, I work in the healthcare space and I looked at them and I said, you know, you're right. Co-pilot probably can do that. But, you know, would you want your loved one to have brain surgery using a kitchen shear? You know, it's gonna work. It's gonna work. But it's not the cleanest way to do it. You know, that's not the right tool for the job. And so, you know, I think that, you know, when you look at over exuberance, you know, there's some kind of level setting and education, you know, because I think, like you said, you know, at those higher levels, which bless their hearts, because it's like, you know, looking at millions and billions of dollars of budget and so many people's livelihoods and whole organization perspectives, you know, you can't afford to get into the nitty gritty. You don't have time, you know, or energy for that matter. So, you have to look at the big picture. But there also, you know, has to be a little bit of education happening about what is the capability of said AI tool. You know, is this AI tool the right tool for what we need to have happen? And the answer to that might be yes, the answer to that might be no. You know, I think that, for me, you know, I will say sometimes, you know, Tim needs to be brought down from the rafters a little bit, you know, because I'm like, hey, I can do this, this, this, and they're like, come on back down, come on back down to earth. You know, but, but that's because we're level setting, you know, my team will be like, let's talk in reality sometimes, you know, we go that way. But yeah, I think the other aspect of that, though, and that's kind of bringing and educating senior leaders in the reality of the situation. But also, you know, the over-exuberance and where this can also come from from senior leaders is that also that capability of AI. We live in a time where AI can provide incredible results. You know, I'll give you an example in the security space. They make tools now for X-ray machines, like they can see at the air.
airport or the entrance of the court house or something. And these tools utilize AI to analyze what that image is being seen and identify potential threats. And so it's like an extra pair of eyes in addition to that human trained operator to potentially see things or components that maybe a human might not pick up on because bad actors are trying to do things that are trying to be evaded by. But AI recognizes and says, hey, this might not be so great. You should probably take another look at this and identifies that operator to make that decision. So like, you have AI in the healthcare space. You have AI's that are helping providers in miraculous and amazing ways to identify ailments or in images or things like that. You're helping AI's or helping folks right and be more efficient in that space as well. So you can see from the senior leaders perspective, we have these tools that are really amazing and innovative. And we want to get in on that because we want to be at the forefront in those organizations. Yeah, it's kind of this use. I mean, this is one of my favorite phrases, but use your powers for good. Like use your human powers to the best of your ability and use your AI powers to the best of your ability. And those two combined seems like should be the ultimate formula for success, if you will. And in that learning and development space, we look at using our powers for good and kind of circling back to, we use our powers for good to make ourselves more efficient so that we can be more present for our learners, you know, and so that we can build better interactions, better courses, possibly faster, possibly cheaper. But we focus more on kind of how the technology helps us and supports us, how we work with that AI as opposed to use with that AI. Yeah, love that. So a lot of what we're talking about is everyone reaching or getting to and maybe it's constantly evolving, but basically our own comfort level with AI kind of knowing when to use it, when to use our human self, when to rely on it, when to make sure we're not over relying on it, not over exuberant, all of those things. So I'm curious about how you and you mentioned you work one on one with a few team members to help them experiment with the new tools. So start small build up from there, like storyline specifically. How do you how do you do that? What do you say first? How do you follow up kind of how do you help somebody who may be even a little resistant to start working with and experimenting, working with the AI? Great question. So usually I honestly I start with something like co-pilot, okay, even before we get into storyline, like if I'm starting with somebody that, you know, they might be very resistant or kind of, you know, haven't really explored. They knew it. It's out there, but they've never really gone to it. They've never really played with it. And I think that that number of people that fit into that category, by the way, is far greater than we might think. You know, I think, you know, those that kind of work with AI and use AI, I think everybody's doing it. But I think there is a very large chunk of people in the L&D space, especially that they know it exists, but they haven't really dabbled with it, you know, which is kind of where that I think that resistance is sitting. But you know, I always start with something like co-pilot or chat GPT. And, you know, I'll say just ask it a question. Ask it something like if you're working with co-pilot, like analyze my emails and tell me where I can be more efficient. Yeah, like we talked about earlier. Yeah. You know, something kind of practical and useful. And it's like the, especially for somebody that's never seen it work before, you know, it's the look of, oh my gosh, this is magic. I was on a call earlier today and we were developing an infographic for a project. And my colleague commented, it's like magic. It's like watching magic, you know, because it's just ironically we use cloud to, to, build the parameters and put the cloud prompt into the co-pilot to develop the infographic. And there we were, you know, and it was just, she's right. It's like watching magic. And so that's where I start is, you know, is, is start with the large language models. And then I'll start with kind of the image aspect of things. I was teaching a class a while back and I was showing kind of, you know, build images and create images and co-pilot, you know, or your AI tool of choice. In this case, it was co-pilot, but, and I had a cardiology person in the back room and they put in just, you know, I want an image of a human art. And highly, highly educated person. And like this image comes up on co-pilot and I could see in there, it's like, it's like, like vein by vein, like, like, ventricle by ventricle, like really looking at the detail. And I mean, I think the mouth hit the floor because it was like, this is 99.9% accurate, like something that that person could use in an education setting, you know, and that saved the time of having to either go out and find a stock image of a heart or, you know, go out and license an image. It was just right there in front of them. That is amazing. And such a good entry point because you're just saying just try something. Just try it. Just see what happens when you try it. Okay, what about Tim? Because obviously you're pretty far along in this AI adoption journey, I would say. But you had to get here somehow. So are there any mistakes you made when you were trying to figure out AI? Like something didn't work the way you expected and what you learned from it? I think, you know, you said something really point it there. I think that AI isn't a place you go to. It's a journey you take. You know, and so, so to answer your question, you know, there's been plenty of times when I've sat down and I've been fighting with chat GBT, you know, and it's like it's literally like I keep, you know, and raise your hand if you've ever done that, you know, it's like we're all, we're all, if you've done it, you've done it. If you know, you know, but part of that journey is learning how to prompt, learning how to talk to the AI, learning how to write your prompts so that you get what you're looking for. I'll quote, I think his name was Justin Schaefer from DevLearn a couple years back. He did one of the keynotes and he said your AI is like your super smart intern that lacks common sense. And I'm like, that has stuck with me ever since that session and, you know, and it absolutely is. But, you know, as far as like, like struggle street, there's just times and I've seen it especially when I've been doing like coding work, you know, I do a lot of work in like power automate and power BI. And I've been trying to do like really complex workflows and formulas and I'm like trying to tell chat GPT, this is what I want. This is what I want. And it just keeps literally giving me the same thing over and over again. And you know, those are the moments when you just have to step up and walk away. You know, but, but I think where I've learned from that again, goes back to what I said earlier, like not all chat, not all tools have made the same. And so, you know, what I've learned from that now is that, you know, for example, chat GPT is great for a lot of things, but cloud is just so much better at the coding stuff. And it took me a while to get there. So now whenever I'm doing any like power BI coding, DAX coding or power automate, you know, formulas of functions, like I'm not going to chat GPT in ways of my time there. I'm going to cloud because I know that I'm going to get a better, better output. Yeah, I think you bring up a good point. And I've, I've made some discoveries myself about which AI tool I like to use for which thing, because which some of them are better than others. So maybe one of the things to try is just try, if you can, try out a number of different AI tools and use, I mean, right now I sometimes will do, I'll go to chat GPT and cloud and I'll ask them both for the same thing. And then I'll see so I can compare because like you, what I found is chat GPT does a pretty good job when I need like marketing copy or something, but cloud does a way better job for me when I need to analyze big, huge chunks of qualitative data, which I often need to do. So I just, there's, I think figure out which AI you want to use for which project or task or whatnot, because they are not all equal. And that is such a great thing to say, because like I'm sitting here thinking, you know, I use chat GPT for kind of like my, I'm going to call like the fun stuff, like the cutesy stuff, you know, I use cloud for the serious business stuff. And then I use Gemini for all like the media stuff. Like, you know, I've done through the Google labs, like the video creation, you know, using their tools. And quite frankly, I mean, I pay $60 a month for all three of them, you know, and that's a utility bill. I look at them as a utility bill, you know, because they're so integral to my daily work, you know, just like you have, you know, your cable provider for your cable, you're an internet provider for your internet, your phone provider for your phone. Like, you know, you have a lot for your business AI, chat GPT for your cutesy AI, Gemini for your media AI, you know, or whatever the breakdown is. And, you know, that was in ways you could look at it. But definitely for sure, everything has its strengths and limits as well. Yeah, I would agree with that. And part of it is just trying them all out and seeing what works when. Okay, there's one question I really want to ask you before we kind of get to our closing pieces because I have a lot of conversations with
learning and development or talent development professionals in a lot of different areas. And the one thing that it seems like a lot of organizations that learning and talent development people are being tapped to teach everyone to use AI. Which, or I mean, I don't, and I think that happens more often even than the increase AI adoption, which scares me because, you know, we can't just train somebody on something and then expect that they're going to use it. So there's the adoption piece and the change management piece too. But you mentioned a one-on-one approach, which we can do when we're working one-on-one with a team member. But now I'm in learning and talent development and I've been told you need to train everybody on AI. And maybe I don't even feel like I'm an expert in AI to begin with. So what do you recommend? How do you respond to this? What's a good way for LND to respond to it at scale? So I'm glad you asked that because ironically and it's timely. I'm working on some ironically executive level AI horses for this exact purpose. And I think one of the key ways to address this, first and foremost, and full disclosure, I'm a data nerd. So, you know, data nerds perk up, you know, is look at the data. Any organizational AI tool, co-pilot is what I'm most familiar with, is going to have data on your usage. And so, like any good, you know, kind of analysis that you would do, you know, look at the data. You know, what is the data story tell you? You know, are people using certain features, you know, highly already? Are we not using any of those features? And let that kind of guide your starting point, I think maybe even before looking at the data, you know, you're going to start with what is the end goal? One of the end goals that I know I'm, I'm nearing dear for with the people that I'm working with is we want to use AI and work with AI specifically so that we can reduce them out of administrative time and allow that time and turn to be geared more towards supporting our team members or supporting, again, in my case, that, you know, the patients we work with, we're getting away from that administrative office lock task type stuff and more back to that human type stuff. And so, you know, to answer the call of what L&D teams might be seeking here is number one, figure out what's the why? Why are we trying to use AI? Why are we trying to incorporate this into our work? What is the output goal here? You know, and then look at the data, you know, where we at, you know, look at the tools that you have available to you, you know, and design a learning path that way. Much like you design any other type of learning, the difference is is that the AI, you know, is a tool that, you know, again, there might be people on any edge of the spectrum, you know, they're either resistant, exuberant, you know, or overreliant, you know, in any category there. So, figure out kind of the why, what are we trying to do? What are we trying to accomplish? You know, look at the data and then, you know, go from there. So, it's like any other project, honestly. Any other any other development, you know, maybe somebody comes to us and they say, we have to do this because everyone else is doing it or we know this is a good idea or we just feel like this is the right thing to do. And we need to steer down to get to, now we want to have the goal that drives us and the data. And so, if we have that overarching goal and the data, that should be what drives our, our process, our programs, our whatever our strategy is from there. Yeah. 100%. And on top of that, I mean, I would even say, as L&D practitioners get started, you know, if you fall into that category of, you know, you haven't used AI or you're worried about AI, you know, just pull up a chat GBT, create a free account, you know, and ask it to write a script for how to make a grilled cheese, explain a video, and just see what it comes up with, you know. And there's great tutorials out there on YouTube and many other L&D creators have done, you know, AI kind of boot camps, if you will, I'm on prompting and things like that. And just just experiment in place. So that way, inevitably, because I do think that those requests, if they're not there now in the L&D departments, like they are going to be coming, you know, and they're probably going to be coming sooner rather than later, you know, so we can get ahead ahead on those requests. Yeah. Love that. Well, Tim, I was going to ask you, where's a good starting point for somebody? But I think you just answered that question. I think I did. Open it up and try it. Just do something a couple simple. And even just, you know, like we were talking about earlier, every tool has its pros and cons. And the Tim Frost recommendation is, you know, ChatGBT is good for that. Like I was saying earlier, it's good for like the scripting or the the kind of text-based art type stuff. Even imagery, I would even kind of fall under ChatGBT, you know, if you need images created, if you need something very technical, you know, look at Claude, coding type things or as you said earlier, like a large data analysis, you know, maybe you have to analyze that data and synthesize that AI data usage data, you know, plug that into Claude. It does very, very well with that video type stuff. You know, you have your Gemini's, you know, the Gemini labs with the V03 models. I found that that's really good stuff out there. Josh Cavaliere has a whole bunch of AI great stuff. I love Josh. He's good people. Shout out. He's been on the podcast before. Okay. He's been here. We can throw that to that episode too if people want a second one to listen to it. But yeah. Beautiful. But he's done good work with that stuff. And so, yeah, just dabble in kind of what you're trying to do and pick your, pick your poison, pick your flavor. Awesome. That's great advice just to get started. And it's amazing because I know, I mean, it wasn't that long ago that I was doing the same thing. Talk about how much things have changed even in a year. And I think that was how I started was all right. I'm just going to act like I have an intern. And I'm just going to start talking to it like it's an intern that is never offended when I tell it to do something or I tell it I don't like something. And you know, I would even take it a step further because like now that you said that, like, I'm thinking back to like the day that I open chat GBT for the first time. Because you know, you hear it. It's like, what is this like? Okay. And it seems like so long ago, but really it was what like two and a half years ago, three years ago, something like that. The most. But it's like, it feels like it's been around forever. But you know, I, the first project I ever used AI on, I created personas, I created characters and backstories and quite frankly, actually very complex medical scenarios. And it blew my mind because it was a, a need that I had. I was like, here's a tool that I can use to try and do this. The alternative was I sit down with a smee for hours and try to interview them and write it myself. And we were talking like maybe 40 different personas that we were trying to create. And I mean, it did it in an hour, you know, of me prompting back and forth. And then this need just had to validate all the data. And I mean, they were surprised. They're like, hey, I did this because it was, I think 99% accurate. And from there, I was hooked. So, so they, I know it sounds cliché. Just sit down and try and get started. But, but find a use case, you know, whether it's a project you're working on or, or, or, you know, something and that's true of any of the type of models that you're looking at. Just plug it in and get started. I mean, I've even asked it, can you do this? Can you do, you know, and then it will say, yes, or no, or what it'll say, here's what I need, or I can say, can you do this? What would you need to do this? Like, you can even ask it, those kinds of questions. Have you ever done the, I don't even know if it was an official challenge? So, I might be totally making this up. But like, take a picture of your pantry and your fridge. Okay. All right. So, if you're curious, if you really want to get, get crazy, take a picture of your pantry, take a picture of your fridge. Plug those images into your AI tool of choice and say, this is what I have available to me. I'd like a recipe for dinner. And see what it comes up with. Okay. I mean, that is one of the things about being a grown-up that nobody warns you about is that you have to figure out what to cook every day of your life forever. So, this is amazing. And we've tried one and it wasn't bad. Okay. Good. Okay. Good. So, look at AI tips and dinner tips. Dinner tips, right? All in one podcast. This is amazing. Well, Tim, this has been a delightful conversation. I can't wait for others to listen to it because I know they're going to find value in it. If somebody wants to reach out and connect with you, learn more about you, what's the best way for them to do that? Beautiful. Well, I am on LinkedIn, of course. So, hit me up on LinkedIn. Feel free to connect with me. Also, feel free on our ATD Central Florida chapter here. Great. So, feel free to look at our chapter and hang out with us on any events. We do a lot of work in this space. So, we'd love to connect with all your listeners. And thank you so much for having me. This has been a delight. Totally was. All right. Awesome. Thanks so much, Tim. Thanks for joining us on this episode of the L&D Must Change podcast. If this episode resonated with you, I've got a couple next steps. First, click subscribe on Apple podcasts. So, wherever you listen to make sure you don't miss future episodes. And while you're there, post a quick review so others can find a benefit from this podcast as well. Next, share this episode and your key takeaways and action steps with others. You can do this in social media. I recommend LinkedIn or simply through conversation. It is amazing how powerful.
A personal share can be in order to both help you make sense of and solidify what you learned and inspire others with ideas on how they can make a change as well. Be sure to tag me and the guest of this episode if you do post on LinkedIn so we can participate in the conversation as well. There really isn't one single person who can improve this profession alone. So let's change L&D together for the better. See you next time!
Podcast Summary
Key Points:
AI adoption in L&D is at a "tri-linked crossroads" of exuberance, resistance, and overreliance, with professionals oscillating between these attitudes.
Exuberance comes from vendors and leaders seeing AI as a cost-saving tool; resistance stems from fear of job replacement and reluctance to learn new tech.
Overreliance occurs when people use AI to replace their own thinking, such as relying on AI-generated notes or content without critical input.
The host shares a personal example of using AI for dietary advice, illustrating how AI enables instant, personalized, iterative learning outside formal structures.
The guest argues that AI is a tool, not a replacement, and that human judgment remains essential—comparing AI's limitations to robots failing at simple tasks.
L&D must shift from reactive, request-driven work to strategic partnership, using AI to advocate for value and tie efforts to business outcomes.
Summary:
In this podcast episode, host Jess Omley and guest Tim Frost explore the current state of AI in L&D, framing it as a "tri-linked crossroads" of three conflicting attitudes: exuberance, resistance, and overreliance. Exuberance is driven by vendors and executives who see AI as a quick fix for cost savings, while resistance arises from fear of job loss and the effort of learning new tools. Overreliance occurs when professionals let AI replace their own critical thinking, such as using AI notes as personal thoughts.
Jess illustrates AI's practical potential by describing how she uses ChatGPT for personalized dietary advice, highlighting how AI enables instant, iterative, and non-judgmental learning—a model she believes should inform workplace L&D. Tim emphasizes that AI is a tool, not a replacement, pointing out that even advanced AI still fails at basic physical tasks. He argues that L&D must use AI to advocate for its strategic value rather than be seen as replaceable.
The episode challenges listeners to reflect on their own attitudes toward AI and how these impact their work, encouraging a balanced approach that leverages AI's efficiency without losing human judgment. The goal is for L&D to evolve from reactive request-takers to proactive strategic partners.
FAQs
It is a conversation dedicated to uncovering changes learning and development professionals must make to solve modern talent challenges, hosted by Jess Omley.
It addresses the clarity and language gap that prevents L&D teams from being strategic partners, showing where a team stands across nine competencies and providing a plan to close gaps.
The three points are exuberance, resistance, and overreliance, with L&D caught in the middle.
She describes it as mostly curiosity paired with a teensy bit of skepticism.
She used ChatGPT as an instant, non-judgmental dietary assistant to get personalized breakfast and lunch ideas based on her goals and restrictions.
He believes L&D needs to use AI tools to advocate for itself and demonstrate organizational value, so it isn’t seen as replaceable.
Chat with AI
Loading...
Pro features
Go deeper with this episode
Unlock creator-grade tools that turn any transcript into show notes and subtitle files.