Join the discussion: Ask Kambria and Jordan questions about AI and learning
Upcoming Episodes: Check out the upcoming Everyday AI Livestream lineup
Connect with Kambria Dumesnil: LinkedIn Profile
Learning and development (L&D) is undergoing a profound transformation, guided by the advancements in Artificial Intelligence (AI). Just like every other industry, L&D is on the cusp of a major change that will radically impact the way businesses train and upskill their workforce. In a recent episode of the Everyday AI podcast, host Jordan Wilson and guest expert Kambria Dumesnil, founder of AI Innovation Lounge, discussed the exciting potential of AI in reimagining the learning experience.
Enhancing Workplace Learning
Traditionally, L&D efforts have relied on methods such as e-learning modules and virtual webinars. However, generative AI is poised to revolutionize the effectiveness and engagement of workplace learning. Dumesnil emphasizes that the key lies in leveraging tools like Chat GPT and GPTs to create more interactive and personalized learning experiences.
Kambria highlights the power of microlearning, where AI simulations and scenarios can be used to enable learners to practice real-life conversations and enhance skills in a low-risk environment. This immersive approach fosters authentic engagement and ownership over the learning process. Imagine sales teams being able to pitch clients, change managers practicing conversations with stakeholders, or leaders improving their communication skills, all with the guidance of AI avatars.
While the potential benefits of AI in L&D are clear, some organizations have been cautious in adopting these technologies. One major barrier is the lack of a coherent generative AI policy within organizations. Without adequate guidelines, there can be hesitancy and fear of making mistakes. However, the risks can be managed by building upon existing L&D models and applying best practices.
Kambria emphasizes that generative AI should not replace human expertise in L&D, but rather augment it. By incorporating their own knowledge and experience into AI prompts, L&D professionals can ensure that the outputs align with adult learning theories and organizational goals. Through testing and continual improvement, organizations can gain confidence in the effectiveness and reliability of AI-powered learning experiences.
Looking ahead, the integration of various AI tools and techniques will be key. It is crucial to move beyond isolated silos of AI implementation and create an integrated strategy that maximizes the impact of AI in L&D. By combining different AI models and applications, organizations can deliver seamless and meaningful learning experiences that cater to the diverse needs of their learners.
The episode emphasizes the immense potential of AI in L&D, specifically in the areas of adaptive e-learning, performance support, and self-directed learning. As AI technology continues to advance, organizations will have even more opportunities to leverage AI in reimagining the learning experience to drive business growth.
AI has the power to transform the learning and development landscape, enabling organizations to tailor their training programs, improve engagement, and empower learners at scale. Business owners and decision makers are encouraged to embrace this evolving technology and explore the possibilities it offers. By partnering with professionals skilled in AI and L&D, organizations can unlock a new era of learning that seamlessly integrates AI and human expertise, ultimately driving growth and success in the digital age.
Topics Covered in This Episode
Jordan Wilson [00:00:17]:
Learning and development is probably going to change pretty soon. Just like about every other industry out there, l and d is something that generative AI is probably going to change a lot, and that's what we're gonna be talking about today on Everyday AI. Welcome. Thanks for joining us. My name is Jordan Wilson. I'm your host. And if you don't know, now you know. Everyday AI is a daily livestream, Podcast and free daily newsletter helping everyday people like you and me not just learn what's going on in the world of AI, but how we can actually leverage it.
Jordan Wilson [00:00:54]:
And we bring on experts from other fields so we can talk about how AI is impacting our day to day, how it's impacting our careers, and how we can use all of these new advancements and techniques and grow our companies and and grow our careers. So, I'm extremely excited about today's show, and a little remix for y'all. It's technically prerecorded. Don't worry because myself and our guests are still gonna be in the comments Responding live. So let me know as we go along. How do you think, the the learning experiences is going to change with AI? Because It's going to change a lot. And if you are looking for your daily news, don't worry. We still have that.
About Kambria and L&D
Jordan Wilson [00:01:33]:
Make sure to go to your everyday AI.com. Sign up for the daily newsletter. We're gonna have all of today's news in there. We never take an actual day off. It's still there. So I'm excited for this conversation. Let's get ready to talk about reimagining the learning experience with AI, And please help me. Welcome to the Everyday AI Show.
Jordan Wilson [00:01:54]:
We have Kambria Dumesnil. She is the founder of AI Innovation Lounge. Cambria, thank you so much for joining us.
Kambria Dumesnil [00:02:02]:
Yeah. Thank you. I'm excited to be here.
Jordan Wilson [00:02:04]:
Alright. Let's do this. Well, hey. Tell everyone real quick just what you do at the AI Innovation Lounge.
Kambria Dumesnil [00:02:10]:
Sure. So I have the AI innovation lounge that I set up, early 2023 once ChatGPT was released, and I really wanted a place to experiment with AI in learning and development. Since then, it started just me on my own seeing theoretically how it could be used, And then I started to work with different companies and integrating it first into workflows because that's the easy quick win getting it To help create your course objectives, your course outlines, and using it to really optimize your process. And from there, it just kind of expanded out once We really realize the capabilities of what it can do to the learning experience and part of the learning flow rather than our workflows. So that's really where the space I'm at now is Reimagining that experience and being able to bring in different AI avatars and different Learning simulations. There's a lot of opportunities there.
Jordan Wilson [00:03:04]:
Yeah. And and, you know, even for those, of our listeners who maybe aren't as familiar, can you just kind of Blaine, you know, what learning and development even is and how it really plays out in different organizations.
Kambria Dumesnil [00:03:17]:
Sure. So with learning and development, I'm focused on the workspace rather than, an educational context. So for us, It depends on the organization, but usually there's an LED team. There can be a generalist who's maybe is an instructional designer, also does the elearning development, Or those can be specific roles, but really what it's about is giving our learners the information and materials they need so that they can do their jobs well and they can perform. And a lot of times what we hear with that in learning and development with instructional designers is we kind of feel like order takers sometimes because the leaders will come to us and They'll say I have this problem. Maybe, my customer service team isn't performing correctly, and we need a 2 day training on this. So that's where we come in. We try to get a little bit of context for that, understand what's going on, see if training really is the answer.
Kambria Dumesnil [00:04:08]:
And then from there, we can build out different e learnings to support that if that's the need. It could be a workshop that we have. It could be a microlearning that we have, maybe just quick video or email, bringing awareness to general principles. So it's really broad, the solutions that we have available. But what's really important for us is being able to dig into the root causes that are going on and really defining what the needs are and what the desired performance even is So that we can work to close those gaps.
Jordan Wilson [00:04:37]:
Yeah. And, you know, pre generative AI, right, how Or or or maybe what are some of the most effective kind of l and d techniques so then we can kind of, set it against what it's gonna be post AI. So let's quickly talk about that. What's more you you know, what was most effective before generative AI came onto the scene? Is it is it just elearning? Is it internal, you know, sub, you know, sneeze. Like like, what is that, you know, that go to for l and d?
Kambria Dumesnil [00:05:07]:
Yeah. I don't know if so. I could say what's really most effective because it's very context specific. But what's really, most common, I would say, Especially with post COVID is elearning and, webinars online, virtual experiences. And for that, we really work with our subject matter experts. So if I'm working with, someone in leadership, I'm not necessarily an expert on being a leader or a supervisor, But I can work with the people who are and get that information from them and then deliver that to our learners in a way that they're gonna be able to take that. They're gonna be able to learn what they need to. They're gonna be able to apply it and get that back to those they're interacting with.
Kambria Dumesnil [00:05:46]:
So if it is a elearning, we have a lot of Different tools we can use within that, like learning scenarios where you go in and you can practice, different scenarios that'll pose, and you can Click on the one that you think is the most appropriate response. But with generative AI now, that opens up a lot more possibilities because it's not just predefined Situations that we're working with. The AI can actually adapt to what it is that our learners responding, what it is that they're saying, and Different things going on there to bring a little bit more meaning to that and more opportunities for application.
Ways GenAI can make learning interactive
Jordan Wilson [00:06:22]:
Absolutely. And, you know, you bring up a good point, that, you know, post COVID, Alan D maybe just it it sounds like it just became a little, less personal and and maybe less engaging as, the virtual aspect and the elearning maybe took a little more and prominence around, you you know, gathering everyone together. With that in mind, how can generative AI, maybe bring some of that interactivity back. Right? Because it is still, you know, mostly, I'm guessing, in front of the computer, but, how can AI help, you know, bring, maybe a little more life or a little more, engagement into that experience?
Kambria Dumesnil [00:07:04]:
Yeah. I think there's a lot of opportunities for engagement. It doesn't have to necessarily be us as a facilitator, but AI can take a little bit of that role from us. So instead of having just an elearning where they're clicking through, we can really take microlearning is a really big one where Sometimes they'll watch a video or sometimes they'll just go and complete a quiz to see how they're doing on something, but we can make that really interactive and engaging and and Meaningful for their experience. So the simulations are one that I work a lot with, but some of the organizations, there was one We had this for a new change manager training, and they needed to be able to practice different conversations with stakeholders. So you have, like, the regional, managers all the way down to line staff. Each of them, when you're having real life conversations with them, they have different needs. They have Different desires, and even within their role, they have different perspectives and different barriers to change that you need to address.
Kambria Dumesnil [00:08:01]:
So Being able to authentically practice some of those conversations with an AI, not text based, but having the actual conversation and hearing the types of questions they might have, Hearing what might make them tick a little bit and how they might react if you say something a certain way. That's where we really get to have some of the that more authentic engagement. And it's not just watching a video. It's not just Clicking through an elearning that we've developed for you. It's you being able to really have ownership and decide what you need to practice, when you need to practice it in In the way that's gonna work best for you.
Reception of AI in L&D
Jordan Wilson [00:08:38]:
You know, I'm curious. How has and and maybe not even in your personal experience, but if you can talk, you know, A little broader, just the l and d space in general. How have they taken to generative AI? Is it something, You know, l and d, all all departments are rushing toward it. Are they cautious? What's been the reception so far? Because it's only been, you know, kind of, and Quote, unquote, mainstream for, you know, little less than a year. So what's the reception been so far?
Kambria Dumesnil [00:09:07]:
I think the reception, it's A little bit different. So if you were on my LinkedIn and you see everything I post, it would seem like everybody's using AI and everybody in learning and development is in that space, but They really aren't yet. I've had conversations with a lot of organizations who are just thinking about it, and they wanna know what could it possibly do for us, What about the risks that it brings? Because those interactive scenarios I talk about, those do bring new risks because you have less control over what the AI is saying and how it's responding. So those are some things to think about, but the ones that are starting to experiment, a lot of them are just within the workflow optimization because that's a piece of it that We can control the most if we're using the AI system to help us with our processes. And then once we get that comfort level, That's where some of the other organizations that I have worked with, they're ready to have the learning experience, reimagined. And they're ready to try new things and experiment with, kind of more lower stakes situations is where we're at right now. We're not creating a safety and health training for a crane operator yet because we're just not there, and it needs to be more than just conversation and text based for that. But there are a lot of opportunities where we can do it, and we can do things that we haven't done before, so it brings more scale to it.
Kambria Dumesnil [00:10:24]:
We're taking smaller trainings that may have been in person. We could only do it with a select group of individuals or not even in person if it was virtual through Zoom. You can now give them that authentic practice, wide scale and bring that to them. So that's kinda what I'm seeing is we're not all there yet, but I think there are still a few organizations that aren't getting there, and they're gonna be on the forefront of this in learning and development. And the professionals in that space too, that's really to their advantage and their career because it's gonna bring more possibilities.
Jordan Wilson [00:10:55]:
You know, Camry, as you kind of talk about the current, scenario and and kind of the adaption rate to to generative AI, You know, there's so many ideas going off in my head and so many, you know, oh, why not? So, you know, maybe why why not? Like, why do you think People aren't rushing more toward this. Is it more, maybe l and d, you know, groups within organizations aren't as Aware of how flexible, generative AI is and its capabilities. But but maybe why do you think or or maybe what are some of those those holdups Do you think?
Kambria Dumesnil [00:11:31]:
I think the biggest barrier starts with the organization themselves because if they don't have a generative AI Policy in place where they haven't really given their blessing to do these kind of things. It's really everyone for themselves, and a lot of Teams and individuals, they're afraid to try it out if their organization hasn't said they could or even if they said they could and not giving guidelines on how to do that. You don't wanna do it wrong. So I think that's one of the major holdups. But beyond that, it's Really understanding what capabilities AI brings and how you confuse that with your way, work in a way that still allows you to be the expert. We don't want AI replacing us, and that's something I think we really need to understand is it's not a replacement for learning and development. But if we don't take charge of it, it could be because it could be somebody else in your organization deciding, I'm the subject matter expert. I'm gonna make a training on this because I now have access all these tools.
Kambria Dumesnil [00:12:26]:
And they might do it in a way that doesn't necessarily align with the adult learning theories and all the background that we bring into that. So if we wait too long to adapt, that's where I think we have some of that risk of it might not be replacement right away, But it's definitely gonna diminish our value as learning and development professionals if we can't provide what's expected of us in some of the organizations.
AI uses cases in learning
Jordan Wilson [00:12:48]:
So so as you kind of balance that, right, you know, having to go through the proper procedures of, You know, maybe getting a generative AI policy, but then also, you know, with somewhat of an urgency of of of wanting, You you know, the background of of l and d teams to be represented in the trainings going forward. Maybe what's what's 1 or 2 Specific areas, that you think are maybe the best use cases or, you you know, maybe even talking about certain Certain tools or certain techniques of of leveraging generative AI in specific use case scenarios. Maybe What are 1 or 2 of those things that you're kind of excited about and you see extreme potential that saying, hey. Y'all, this this could work. Let's do this.
Kambria Dumesnil [00:13:35]:
I think the biggest area I see is with microlearning specifically just because it is so static. When we look at it, we have our videos and, we might have an elearning, but if we can bring authentic scenarios and, different simulations, especially around conversations, I think the conversation topic is going to be a really big use case for us. So if that's leaders talking to their employees and having 1 on 1 conversations growth and development, where it could be, like I mentioned, with change managers having those conversations. It could be sales teams, pitching a client. There's a lot of different and Opportunities out there, and these scenarios are so easy to build once we have the knowledge of what it is and And what it means to perform well. And we have the persona. We know the audience that they're going to be having that conversation with. That Makes it pretty quick to build out those different simulations, and there's a lot of tools available.
Kambria Dumesnil [00:14:33]:
So the one I did with change managers was with Inworld AI, but We also have GPTs. We also have if we wanted to do, text based, there's a lot more opportunities. But the voice ones, I think really in world AI And GPTs are where it's at right now for having conversations and dialogue back and forth, but that's also going to expand too as other tools take off, and we have more options available with.
Using custom GPTs to learn
Jordan Wilson [00:14:57]:
Yeah. And and that's that's even where my head was going. Well, 2 different places even just within chat gbt. It's obviously something that We talk about a lot on this show, but, you know, Cambria, you mentioned GPTs. So, you know, if people aren't familiar with that, it's It's a recently, released feature from OpenAI, but, it allows people without really having to know how to code, But to create a very custom, version of chat GPT, essentially, a very custom chatbot, for a very specific purpose, very quickly. We're talking about in minutes, something highly personalized, highly specialized. How do you see those GPTs, being used in l and d in the near future, or Or do you think there will also be friction toward using those as well?
Kambria Dumesnil [00:15:47]:
I think I see them being used only within l and d teams that have already adopted AI within their workflows. If they're not at that stage yet, you're probably not ready to be looking at GPTs for your learners, but it could be something that might help you with your workflow. For those that are experimenting with it in their own workflows and they're trying out different things with their learners, that brings a lot more opportunities for A lot of different things. There's performance support, being able to upload your manuals. Instead of us manually creating a job aid, they can just Chat with the GPT and ask it questions about the process. As far as I know, I haven't seen image retrieval, but That doesn't mean it's not gonna be there soon or you have Dante AI that does have that. So you could take a look at all the different tools out there and see what works for it. And then for the conversation piece, you can't do that as I I know of within web browsing for chat gpt, but you can on the app.
Kambria Dumesnil [00:16:41]:
So the app itself brings new capabilities and new use cases. And as you start to experiment with some of those, you'll find ones that are work specific to your organization Based on what your learners need and the different situations that they're going to be encountering.
Jordan Wilson [00:16:56]:
Yeah. And that and that that was kind of even my next question getting to that, You know, even with on the or within the chat g p t app, the voice features and and being able to literally have a conversation With a large language model, like that, you you know, if you're able to on the front end as an l and d department or as a highly skilled, you know, individuals such as yourself, you you know, can do that. What kind of value is there there where you can go through and practice With your your actual voice and hearing back from an actual voice, you know, within chat GPT, What are the applications for that? Is that helpful? Is it potentially harmful if you don't have enough Control over what the, I guess, what the output from, Chat gpt and and and and that, voice feature might be.
Kambria Dumesnil [00:17:48]:
I think practice conversations are probably the biggest l and d use case for that base voice feature. And the risk with that really are minimized because Conversations themselves with individuals are unpredictable. So even if ChatGPT is not responding how the LED professional designed the prompt or created the GPT for that, It's okay because your learner could very well encounter that, and they might need to be prepared. Now you don't want it to go completely off the rails, but It's pretty low risk for that happening. So that's where, I had mentioned before we could use that with practicing 1 on 1 conversations with different individuals Or having, the change manager conversations. So there's a lot of different things with that for sales, just across the board. And I think that that's gonna open up a lot of opportunities for us in learning and development, but we can also think from our learners' perspective. We talk about and Self directed learning a lot and having resources out there for them.
Kambria Dumesnil [00:18:45]:
So we don't need to necessarily be the ones creating the GPTs. We need to empower our learners, Build AI fluency within the organization, and then that'll enable them to find their own use cases. And maybe they have a upcoming conversation with somebody that Doesn't happen often, but they wanna practice for. And they can really just go in and build their own custom GPT just for that one time. And they can use that because it's so quick. It doesn't require us to do it for them, and it's gonna be a lot more tailored to what they need.
Future of GenAI and L&D
Jordan Wilson [00:19:15]:
You You know, Kimber, you you you we talked about some of the kind of, lower hanging fruit, you know, with generative AI, some some very practical and and actionable, you you know, options or tools for better learning and development. But maybe as you look a little further ahead, you you know, past some of those things that are technically available right now, What are some areas that that you're looking at, a little further down the line that you're really excited about when it comes to generative AI and how it may, In the longer term, impact the l and d space.
Kambria Dumesnil [00:19:51]:
I think the main thing I'm looking at long term is How we tie all these things together. So we don't wanna just have things in silos where we're building a GPT here. Maybe we have and elearning integration through the API, which is something I'm really excited about where we can have adaptive scenarios within that elearning for us. Usually, we build that in a storyline, and there's a lot of individuals in that space working through the API on that. And the Turbo o four. It's gonna lower the cost a lot for that, and it's gonna make it a lot more feasible. So, really, it's how we bring all these things together To make the overall learning experience meaningful, and it's not just something we do here or there. It's an integrated strategy that Accounts for all the different opportunities and puts them together in a way that makes sense for the learner and is not gonna overwhelm them.
Kambria Dumesnil [00:20:39]:
Because the last thing we want is to have An abundance of content out there and them not knowing what's gonna work for them. We don't need content overload or more out there because that's something we already struggle with.
Advice on using GenAI in L&D
Jordan Wilson [00:20:53]:
Yeah. No. That's that's a great point. And, you you know, even speaking of that, let's let's talk to that side of it because, I even see this personally. If you don't put enough care and attention into just content. Right? Like, specifically with GPT, The outputs are gonna be pretty bad. Right? So, you know, how can you get around that? Because I'm I'm sure as, you know, l and d kinda departments, you know, start to experiment with this. Maybe they're not very well trained in, you know, GPT and generative AI as you are.
Jordan Wilson [00:21:25]:
So so what are maybe some, you know, best practices or pieces of advice that you have for people that maybe do wanna jump in, and and use some of this generative AI, but maybe they don't know enough. And it's almost like it's almost too easy Or so easy that someone inexperienced can go in there and make some mistakes. So what are some, you know, pieces of advice that you have for those types of people?
Kambria Dumesnil [00:21:51]:
For us in learning and development, I would say just to start with 1 tool, and that's going to g b chap GPT. I think it has the multimodal capabilities. If you're really looking to get into AI, that's pretty much the end all be all. And there's other things we can use here or there. But we also need to make sure when we're doing our prompts. We don't need to be prompt engineers or learn all these super specific techniques. They can help us. But what's more important than that is that we make sure we bring Our expertise and our learning and development models into those prompts.
Kambria Dumesnil [00:22:21]:
So if we're looking at designing Training, we wanna look at maybe Gagne's 9 events of instruction and how that might fit into their being able to direct the AI, trying to treat it as our learner. And when we do that, we could tell it how to do our processes, how we would interact with the learner if it's a chatbot that we're creating or What it means to be a good tutor or facilitator and being able to bring that knowledge and expertise into that is what's gonna really gonna make the difference. And it starts with testing. That's the big thing. When you create your prompt, you put that out there, make sure you're testing it. Your subject matter experts testing it, and you can pilot it with learners too to get that feedback and then continue to make improvements on that.
Kambria's final takeaway
Jordan Wilson [00:23:06]:
There's been so much valuable information in in this conversation, you know, as we talk about, you you know, reimagining the learning experience with generative AI. You know, Kimbra, as we wrap up here, you know, because we've been everywhere. We've talked about the risks and challenges, you know, some short term wins, where the space is kind of going in the long term. We talked about prompting. What's your one Kind of takeaway that you want people to hear that are maybe in the l and d space or maybe for, you you know, a a business owner that that needs maybe they haven't really Made l and d a focus of their company because they haven't been able to, but now it's more affordable and accessible. Maybe what's that one, big takeaway that you want people, who are interested in this, to take from this conversation.
Kambria Dumesnil [00:23:53]:
I would say just, like, Main takeaway is not to be afraid of AI taking your job and recognizing, like, for those small businesses who don't have an l and d team, The new opportunities it can bring for you because that's really the main use case for AI as a whole for learning and development is being able to do new things, doing them in a different way, and doing them better. So that's kind of where that mindset needs to come in. And then just starting with that 1 tool, one use case, Taking it 1 at a time. Don't try to do everything all at once. But if you have a single use case, a single chatbot that you wanna try out, Don't be afraid to do that, and you can learn and adapt on the way. That's where AI is really valuable is being able to adapt. So Your responses and what you tell it and learning and improving with you.
Jordan Wilson [00:24:42]:
Hey. If you're in l and d or if you're not, Kambriajust dropped So much knowledge, so much great advice for you to get started and improve your l and d with AI. Cambria, thank you so much for joining the Everyday AI Show. We really Appreciate it.
Kambria Dumesnil [00:24:57]:
Yeah. Thanks for having me.
Jordan Wilson [00:24:59]:
Hey. And as a reminder, there's always more. There's always more. We never stop here. Make sure you go to your everyday AI.com. Sign up for that free daily newsletter. We're gonna be breaking down today's conversation With Kambriain much, much more detail as well as get your latest dose of AI news and everything else. So thank you for joining us, and we hope to see you back for more everyday AI.
Jordan Wilson [00:25:20]: