What AI Can Do For Online Learning Simulations

Dave McCool, founder and CEO of Muzzy Lane, joined me to discuss the role and potential of AI in creating dynamic, role-play simulations for online learning. Dave shared the journey of Muzzy Lane, from its early days developing history games for schools to its current focus on enabling educators to easily build their own customizable, auto-graded simulations across more than 100 higher-education course areas. Our conversation highlights how recent advances in AI have transformed the process of authoring simulations. It’s now much faster, more accessible, and more engaging for both instructors and learners.

I highly recommend you don’t just read or listen to our conversation; watch it because Dave gave a live demo of the use of AI to create simulations. Show don’t tell, as the saying goes.

Michael Horn

Welcome to the Future of Education. I'm Michael Horn and you are joining the show where we are dedicated to creating a world in which all individuals can build their passions, fulfill their potential and live lives of purpose. And to help us think through that, today we have a special guest, Dave McCool. He's the founder and currently the CEO at Muzzy Lane, which basically creates dynamic role play simulations in a variety of fields. We're going to learn a lot more about it, but simulations have been an area that I've been very interested in for a long time in the world of digital, online learning and so forth for its ability to frankly create more real world, real life learning experiences for individuals. Dave, welcome. So good to see you and thanks for joining us.

Dave McCool

Thanks for having me here today, Michael.

Michael Horn

Yeah, so let's dive in maybe, because I think simulations have taken almost like a, you know, level up, if you will, from video game land. Right. With AI over the past few years. But let's go back a little bit further than that. Just the founding of Muzzy Lane. What was the big idea behind it? How's the company evolved? Who are sort of, you know, who do you serve right now? Where are these simulations and where are they hitting learners at the moment?

Dave McCool

Perfect. Yeah. So the big idea is a great place to start. So I'm a software engineer by background and so when we started Muzzy Lane in early 2000s, we had an interest in games and simulations, looked kind of at the landscape and saw education. More digital transformation was happening in education and just felt as a software engineer is like, this is a great place for games and simulations. It's a great place for deeper, more complex interactive software that can really get at better learning experiences and better assessment experiences. So that was really the big idea. 2002 was a very different time than today.

So, you know, we went through a lot of iterations trying to find the right formula. You know, the first half of the company's life was really, we created and released history games into high schools and colleges called Making History. People who use them, love them. Not as many people use them as we had hoped in education at the time, but that series lives on in the commercial space on Steam, so it's still available today. Did a lot of work with publishers. Really kind of most of our work's been in higher ed over the course of the company's Life. A little K12 and a little workforce. But I think for us the big turning point for us was 2014 and 15, the Gates foundation, they gave us a research grant because they'd made a lot of investments in games and simulations for learning and hadn't been getting the results they were hoping for.

So they wanted us to study the market, which was interesting for us as sort of startup software people, to do a research project.

Michael Horn

Yeah, I was about to say, how did that, how did that land for y' all?

Dave McCool

It was good. We. It was a period. A colleague of mine, Connor Ryan, came in and was running the company at the time while I was running the tech transformation we were undergoing. So he had a good experience that we didn't with that. And so we. It was a really great experience.

We produced a 40-page report which is still available on our website.

Michael Horn

What did you learn? Like, what were the headlines from that report?

Dave McCool

I still remember the headlines today. And it was because they were, they were humbling really for us. It was high awareness of games and simulations among instructors and administrators in university, especially online. But all sorts of logistical challenges with what they were being offered. Things were generally too big, too inflexible. Don't disrupt my course became the thing we heard most commonly. We see the value in what you're doing. I'm not rebuilding my whole course around.

Michael Horn

You, but I wanted to fit in as a module as opposed to. To I have to rethink 10 of the 13 weeks or whatever.

Dave McCool

Right. So it still has to be a scale that fits with my course that fits with my learning objectives, like time. Like I don't have enough time to spend on that particular part of my courses that this game is going to require for me. And they wanted to control the content. They didn't want to have to go back to a developer constantly to get changes made and updates made. And then kind of the basic ones were it has to run any device in a browser, on a phone, including had to integrate to the LMS, which meant LTI integration. And it had to meet WCAG 2.0 AA Accessibility Now 2.2, which again, these were not things people. We came kind of from the serious games and the games for learning side and we were kind of like, oh, games are great and you're going to adjust to get those benefits.

So anyway, we were doing a tech transition at the time anyway, that's when the authoring tools came out and we said, let's just go all the way down to we're going to start putting these tools in the hands of our partners, teach them how to use them, make it as easy as possible. And that's really been the business ever since.

Michael Horn

Super. And so what higher ed programs specifically? Like, when you're talking about the simulations, where have you found the most traction? Are we talking like nursing and it's almost VR like, or are we talking negotiations? Like, there's a pretty big range of what these can look like. Where, where have you found the most traction?

Dave McCool

Yeah, I think so. We, when we thought about the tools, we thought about how to help people build their own simulations for learning, we thought a lot about ROI, because when you were building specific games, like we built for principles of Marketing, we built for intro Spanish, we built for operations management, but you're building one big product for one course area and it's hard to make that money back. It's hard to make that work financially. So what we wanted to do is build templates that were as broadly applicable as possible. So we wanted to make content forms that you could then fill in with what you needed for your course area and hopefully those would travel pretty broadly. So we're in over 100 course areas in higher ed. Business is probably the number one by usage overall. But a lot of nursing, a lot of medical fields like medical assisting, a lot of the humanities as well.

So we're in language learning, we're in sociology, psychology, crisis counseling, social work, all those kinds of places. And the role play format of the multiples that we have really has become the most popular one just because it does travel really well and it really fits a lot of what people try to do.

Adapting Games for Education

Michael Horn

And so the, and it sounds like the secret became, okay, we're going to give up our adherence to what a serious game needs to look like or what the ideal version of it is. I'm sort of reminded of the Greg Toppo book, I think the game believes in you, right, where he sort of said there's this clash of like these very big learning goals that serious game designers have as they enter the education space. And like, these are not your words, but these are his, as I recall them, was like, you know, sometimes schools can be a little more transactional in the knowledge and skills that they're trying to get across to students. And so, you know, you may lose sort of the organic nature, right of a game or simulation when put into the container of a class. But it sounds like you were able to take the market's message, make that change and, and give more power to educators to create these role plays. Am I understanding that correctly?

Dave McCool

Yeah, you are. And it's funny because you're I'm. I'm flashing back to all those. Yeah, yeah, and Kurt squire and Jane McGonagall and like all the people who initially sort of had this movement. And a lot of the challenges were created by this clash of cultures between what is a game for a lot of people? Game is voluntary. It has to be fun. What is learning? Some people learning is very different from that.

We're trying to figure out how to fit into the learning environment, push them. Like, we didn't want them to just not change anything because that's not helpful, but. But not impose criteria that are incompatible with what their goals were.

Michael Horn

So talk to us then about, you know, how that's evolved in terms of the product and how do educators create simulations fit for their course now? How do you help them? What, what, what do those authoring tools look like? It seems like a lot of variables and still could be a lot of work if not done well.

Dave McCool

Which is a very good point. So, yeah, so from fall of 2015 was when the platform came out until a couple of years ago, it was a pretty, just sort of a steady, steady workflow for us. It's like, how do we keep making these tools better? We did do custom tools for some people, partners, you know, for example, Western governors early on was, hey, we want to do a pandemic simulation for our masters in healthcare administration, where you run a trauma center during a pandemic. Ironic. That was 2015.

Michael Horn

I was gonna say that turned out to be prescient, sadly.

Dave McCool

But yeah, yeah, but really mostly just finding, you know, great fielding input from customers. How can we make this better? How can make the learner experience more engaging? How do we give you more flexibility to meet your goals? I think for us a big thing is auto assessment was always a big requirement from the market for us. So the tension for us was always, how engaging can I make this and still auto grade the experience? Because a lot of times you're working with online universities, they don't have the ability to grade open response type things. So that was sort of the challenge. And I think that was really just steadily grew and got up to about a million students a year being served. So it was nice. Again, like I said, mostly in higher ed. And then two years ago, really AI, it came out before that, but we were not sure what to do until about two years ago.

And that's really been the newest phase, which has really been. That's been pretty exciting.

AI: More than Just a Chatbot

Michael Horn

All right, so let's talk about that enabler in terms of like, what does AI allow you to do that you couldn't do before? Allow the people that you're working with, your clients, customers to do that they couldn't do before.

Dave McCool

So two really big things. So when we first, when ChatGPT sort of burst onto the scene, all of us, I think, like everybody else were like, oh, wow, this is really, really great, or it's going to make it so that no one needs what we do anymore. We weren't sure which. And the first few, first year or so of looking at it, we thought, well, everything felt like a chatbot, right? Everything sort of felt like, where do we put a chatbot? And we were like, well, we could put it into a smart chat, a roleplay sim, or we could put it in the corner and you could ask it when you got confused, but nothing felt quite right. And I think what, what really struck us a couple years ago, maybe more, 18 months, was instead of thinking about AI as a chatbot that we could put somewhere, we thought about AI as a resource that we could use. So what's AI? It's a big knowledge base of maybe all the world's data, I'm not sure that you can ask questions of and get responses. And we said, well, how could we, starting from roleplay, simulation, what we do, how could that make what we do better? And we really came out with two use cases that we immediately dove in on.

One is we said, what if we could teach it how to make simulations in our format and then provide that to our customers? So, you know, we and everyone else is already using ChatGPT to brainstorm scenarios and stuff. And you're doing a lot of copying and pasting from your chatbot into your tool, whatever tool you're making your content in. We said, let's move, remove that step. Let's just teach it directly how to make our stuff. We're AWS hosted, so we are able to integrate it directly into the platform. And just say, you just start talking to it and you say, I want to build a scenario. Here are my learning objectives, or here's some source material to start from. Walk me through a process that gets me to a fully testable scenario.

So that was one and that's the market has loved that.

Michael Horn

Yeah, I would imagine. I want to come back to that in a moment. Keep going.

Enhancing Learner Experience with AI

Dave McCool

Yeah. The first full version of that came out over Christmas, over the holiday of 2024. And the other was how do we make the learner experience better with this tool? And that was a little more challenging at first and then we finally just said what's the simplest thing we can do? It's like, well we have a, the roleplay simulation has a variety of question types. The usual suspects, multiple choice, fill in the blank, categorize. We said, what if we had an open response question type where instead of using multiple choice in a scenario you've analyzed data, you're talking to a coworker, they ask you a question and you just have to answer. And then what we can do is then use the AI as sort of like a fuzzy logic natural language processor back end to say hey. Michael asked David this, David said this back.

Here's the rubric here. The learning objectives we're tracking. Give me what Michael says back to David, grade David's response against the rubric and give me a rationale for the grading. And so that came out at the same time and that's been really cool too because now you can be much more authentic. Like one of our university partners said, I saw your journey before. It's really interesting. But it was sort of selected response. I saw open response and I said I'll do that changes everything because now it's a constructed response.

Now I, in a scalable auto assessed way can respond in my own words. I can, you know, put my mental model out there and you can, you can, you know, grade that well.

Michael Horn

And it's interesting on that second one. In some ways that also brings it more to the serious games origins and that there's something more authentic I would think about it that's possible. Maybe I'm making the wrong leap there.

Dave McCool

But no, I think that's what we talk about. It's more engaging and it's more authentic. And one thing we did along the way is also we added text to speech and speech to text so that you could actually speak your response. You don't have to type it anymore.

Michael Horn

Yeah. So I want to come back to the first one, but one more question. Can you use AI also like, or maybe this is a couple years off to start building, you know, graphics or like, you know, things that are more immersive or engaging simulations as well and maybe make, you know, there's this been this world of virtual reality that's been five years away for 20 years now. But maybe it becomes more possible because the authoring of these environments really goes down in terms of cost and time. What's your take on that?

Dave McCool

I completely agree. We in the authoring era have really backed off from VR, AR and 3D work which we had done in Our previous custom days, partly because our guardrails were authorable by non technical subject matter experts and accessibility. So those were so kept in that, in that box. We're now doing a really good job of generating with AI all the stuff in that box. What's about to come out from us next is generating images. So there's a lot of images and multimedia in our things today. Building those. And then our next thing is to look at, okay, can we build a 3D scene in, in AI, can we rig 3D characters? Can we create characters that can walk around and interact with each other? So I think that's next.

But that's like we've tried to be really disciplined about staying in like here's what we do really well. 3D and VR definitely has a place. The cost of production and the complexities of deployment mean there are places that's not well suited. So we sort of so far staying where we are.

Michael Horn

No, that makes sense. And so then this ties back into the number one use case around enabling the authoring right to explode in some ways along your guidelines. And I assume that's something I want to let you elaborate on a little bit because I assume anyone can create a simulation, but to do it in line with instructional goals, learning design, not overloading the learner, accessibility, et cetera. I assume there's a bunch of guardrails you all have created over time or rules by which content is created that you know, you don't want a faculty member or if it's at Western Governors, right. Like a centralized instructional designer to have to deal with all that complexity. You almost want the AI to just hard code that in some sense.

Designing for Balance and Structure

Dave McCool

Yeah. It's funny, most of our programming on that part of the product in the last year has been prompt engineering and system file work. And you know what we realized immediately we do not want, we didn't just want to put an AI interface in front of our users like you said and ask them to try to make this. It would not, that would not have gone well. And I think what we constantly, the balance we're trying to find is between how much to let the AI go versus how much to keep it structured. Because like, especially with universities who've reacted really positively to this, a big part of it is they can get the thing they need out of this easily and quickly. And before they could either spend a lot of time and money to get the thing they needed or they could get the thing they didn't really need quickly.

So it's trying to find that middle ground of like, give me your learning objectives, give me your goals, give me your activity purpose. Like is it a formative assessment? Is it a homework assignment? And then I'll guide you through this sort of, we call them playbooks, this playbook process of saying here, here are the steps to get to something usable and here are the ways in which you can kind of vary along the way.

Michael Horn

Last question on this, because you've talked a few times about using it for instruction, but also the assessment inherent in a simulation. Are all the simulations that your customers, universities creating, do they have both of those goals or some, you know, inherently assessment based and some are inherently instructional? Maybe formative assessment based? Like what, how does that work? How are simulations used? Maybe talk a little bit about the use cases of simulations within different courses in terms of their instructional versus assessment purposes. And can, can an assessment, or excuse me, can a simulation serve both?

Dave McCool

Yeah, to answer the last one first, yes, it can serve both. And I think for our entire company history, the challenge has been like, when are we teaching? When are we assessing? Like games? You know, the big picture promise of games for education is that they instruct and teach at the same time. But that isn't always easy to put into an environment that has other stuff happening. So our customers who do all those things, you know, McGraw Hill is a big partner, they do a lot of homework assignment work. So you know, it's a graded homework assignment, but it's not really an assessment in that sense. We have partners who use them as formative assessments, as checks for understanding. So like, you know, the end of each section or module, you might, you might take one that they're using to figure out if you've moved on and you understand or not. And then I think to the other extreme, we've got Education Design Lab, which is a partner in the durable skills space, is using it actually as summative capstone assessment to award credentials.

Michael Horn

Oh, that's interesting. And to show durable skills. It's because if you're, I'm going to make this up a little bit. But perseverance or something like that is one thing. If you can show it in a, you know, a self assessment which is extremely unreliable, but if you can actually exhibit that behavior in a situation where you suffer a setback in a simulation, maybe you can exhibit growth over time on that. Is that, is that sort of, or.

Dave McCool

You know, critical thinking is a good example that's easy to explain to people. You know, you, you, you role play someone in a job who's analyzing data, working with co workers and trying to apply critical thinking techniques to solve problems. Like you're, you're presenting these challenges. You're the produce manager at the grocery store. They want to put a kiosk in your section. Should you do it or not? You have to analyze data, interact with stakeholders and then there are skills, you're exhibiting things like distinguishing facts from opinions, providing thoughtful analysis, identifying core issues.

Michael Horn

So yeah, and so within different subjects, I mean you could in some ways simulations, particularly with the scalability now of AI, I imagine could actually uplevel the quality of assessments trying to get at these things from the multiple choice sort of short answer land to something much more immersive, context based, etc.

Dave McCool

Yeah, yeah, because we've got, I think we've got kind of the uncanny divide right now between how good can an auto graded assessment be or how authentic versus how authentic can a human administered assessment be? There's a big gap there today and we're always been on the side of the auto created, pushing it towards the human. And I think AI really helps you move that bar quite a bit.

Michael Horn

That's really interesting. All right, I know you're able to actually show maybe something, but I would love to show our audience some of what this looks like on the ground so that they can really feel what these simulations are.

Dave McCool

So my plan here is to do sort of the cooking show version of demoing where I'll show you the first few steps in the process of building something and then I'll jump to something that was built with that same process worked all the way through. I'm in our tool right now. This is an activity. So footwear impressions that I created earlier today, I'll go into the editor. So what I'm going to do here is basically this is the use case of creating a simulation, homework assignment simulation from a lesson plan. Okay, so in this case I had previously uploaded one of our favorite lesson plans of late, which is this Texas,, so the Texas CTE standards has a course called Forensic Science and there's a lesson called Footwear Impression. So doing footwear impression and analysis is part of doing Forensic Science.

So I've uploaded that lesson plan previously to the call starting. I can come here and I can say this will be a homework assignment. And then what we have is this eight currently this playbook called Create Activity and it's an eight step process. You can see them on screen right now. You could start at step four if you already had learning objectives and you didn't need those extracted from source material. So some people come there. But let's just sort of start. We'll go,

Step one and what we do at each step is we pre populate the prompt which then is combined with all the system text work we've done under the hood to get back the result for this step. So in this case it's going to analyze that document and suggest the learning objectives that my activity should have. And people do have the ability to sort of edit and play around with these as they go along.

Michael Horn

It's the cooking show thing. I'm supposed to say something witty here.

Dave McCool

Right.

Michael Horn

And pops up with the. Yeah.

Structuring Unstructured Data

Dave McCool

And I think what's interesting about this is, you know, someone said at ASU + GSV last week or a week before. This idea of AI helps take unstructured data and structure it. And this has been kind of an amazing realization for us as we've gone through this process is that you, you talk to the AI in natural language and then it does sort of software program things. And so it makes sort of the trying different things, experimenting getting to your goal a lot easier because you don't have to. There's no. You just try it. Right. So I can say, all right.

So there's what it suggested. Evidence, classification, characteristic analysis. So there are the ellos it suggested. If I go to step two. Okay, gradient learning objectives, let's say so.

Michael Horn

You can carve out an exception.

Dave McCool

I can say sometimes it'll suggest something that doesn't work well in the simulation format, accept evidence, documentation. So I say go. So it will go through those, it's going to add those to my activity. So we can say agree. All right, there it is. If I pop over here to the objectives. There they are.

Michael Horn

Oh wow.

Dave McCool

So like this already would have been a lot more work in time for someone one of our customers existing. And then what we do is we go to step three.

Michael Horn

In some ways this makes it way more self-serve. You all are doing probably less on the back end to help them through.

Dave McCool

Yeah, one of the realizations that hit us last year when, when people started actually using this in our customer setting. Actually now some of those activities are live to students was we realized that they never had to learn the tool. They never left this page. They just worked through these steps, played the activity, maybe made some changes which you can do here as well and then they were done.

Michael Horn

So does that lower cost increase output? Yes, yes, yes.

Dave McCool

Yeah, it's lower cost, it's faster output. I think the most important thing that we've noticed is it takes it from not possible to possible. Like we always say, better, stronger, faster. But we're really in not possible to possible.

Michael Horn

And so that's interesting because you had mentioned non consumption to me when we were trading notes about this. It's almost like people who would never have ventured into this space all of a sudden, hey, I can do something and get it, get it up and running.

Dave McCool

Yeah. I mean we have thousands of universities using our product indirectly through partners who've built for them. But it was always, it was always more challenging to get universities, except say if you're Western governors to take on that task themselves. They are building lots of courses. But they always felt the simulation hurdle was a little too high. And this has really changed that for them. So I ran step three, which was basically to create a story and some characters. So if we go over here and look, we'll see now we have characters that have been created.

So Dr. Maya Rodriguez. And then we get these personalities which are really cool because this is how the AI knows how to, how to speak in this character's voice. And I could edit these if I wanted to, but I just won't. Sure. For time.

Michael Horn

Sure. Yeah.

Dave McCool

Now just do one more step before I jump over to the other side. So now this is kind of the build out a story. So I could say build up the story, blah blah blah. I want to modify that. I'll just let it go. And then, you know, five would be create all the materials, create all the data that's going to support this story so that I can ask you challenging questions. Six would be create the questions. Seven is create the activity feedback.

And eight would be create the intro screen. And now we have a full activity.

Michael Horn

Here. It's generating right now. And so we're gonna. And then you're gonna jump into that activity now.

Dave McCool

Jump into that. The one that I built earlier. Yeah.

Michael Horn

Perfect.

Dave McCool

So we can see this. So city crime labs, unusually busy, string of high end home burglaries. That's everyone on edge.

Michael Horn

Early premise. Yep.

Dave McCool

Right. So there you go. So that's the story. So let's pop up one that was already made and see what the student would see.

Michael Horn

I'm just imagining Carmen San Diego back in the day could have been created in, in five minutes. Yeah.

Dave McCool

It's crazy. Yeah. I mean, so you saw that like I just did half the process in about five minutes. It probably would take another five to the other four steps. Yeah, this one I added a few images to. So I Added another maybe five or ten minutes after that process. But in this one, I told it I want it to be in an art museum. So here we go.

There's a priceless Botticelli sketch has vanished. Your expertise is needed. Here's what you're going to do. Here's your role. I'll come in, set my character.

Michael Horn

Yep.

Dave McCool

Here are the people I'm going to work with.

Michael Horn

Wow.

Dave McCool

Alarm at the museum. Pricessles Botticelli has been vanished. I come up with Dr. Rodriguez to see what's going on. Okay. So we come in.

Michael Horn

And so this could be also the speech. It could be a variety of modalities to interact with it as the learner.

Dave McCool

Yes. Yep, yep. So, yeah, they can voice or not voice. Authors have that choice. And we have as well. I'll turn it off for just a second. Sure. Yeah.

AI as Time-Saving Collaborator

Dave McCool

So. So this data you're seeing on the right here, all AI generated. Created the whole story, created all this data for you, all this dialogue. So I think two things about that. One is that it saves a tremendous amount of time versus an instructional designer or subject matter expert having to do all that, but also provides a collaborator. And, you know, we all work better when we have someone to talk to about the thing we're working on. This actually works pretty well as someone to talk to when you're working on this, which is interesting. Not something we had really thought of beforehand.

I'll just show you what the open response.

Michael Horn

Yeah, let's see it. It's interesting. While we're waiting to come up, I think of all the times I've seen stuff like this that has been super engaging because of the storyline, but how custom and expensive it's felt.

Dave McCool

Yeah. I mean, just a use case here is like, if you really love this pedagogical experience, one of the use cases in the AI is you can say, move this to an auto body shop, move this to a mall, for example, and it can do that. But let's see here. Let's say I think blood seems like a good thing. I might say I think the bloody footprints will be the most useful because they might show us the perpetrator's blood. That was my thought, actually. That's what I thought the right answer might have been when I did this the first time. It turns out it's not.

Michael Horn

I was going to say, are you. What's your criminal forensics background? Interesting. Okay, Right.

Dave McCool

So what's interesting is. So that what Maya just said there was completely generated off of what I said.

Michael Horn

Yep.

Dave McCool

And then the grade, which I got, which I think I got, oh, I got zero out of four. I didn't do very well at all. But then here's the kind of like, hey, here's why you got the score you got. So I think it's really important that learners understand that and also that they can challenge it. So when this is delivered as part of an LMS assignment, there's a challenge opportunity for them to say back to the teacher and the author, hey, I think I should have got...

Michael Horn

This is how I was thinking about it and blah, blah, blah.

Dave McCool

Yep, here's what I was thinking and here's why that should be good. Yeah, so that's, that's the basic experience. You know, you can see over here we've got score categories that are being tracked. There's a feedback screen at the end which gives them a bunch of feedback which we're now generating with AI. But again, just trying to keep making it easier to create these very sort of targeted, authentic experiences for learners.

Reimagining Online Course Engagement

Michael Horn

I'm just blown away. It's interesting. I remember the first online course I ever had the chance to take was I'm going to date myself, 20 some odd years ago when I was, you know, enrolling in HBS. And I don't know if you guys actually helped create it, but it was a stats class and it was all simulation based and I'm just imagining like I thought, geez, if everything could be like this, it'd be amazing. But the other thing that occurs to me is you can also vary the format through this much more easily as well so that you keep it a little bit unexpected, you keep me on my toes. I don't, you know, one of the problems sometimes I think with online courses is yes, it's at my pace and path, but it becomes a little monotonous and predictable perhaps and I sort of the thrill of novelty wears off. There's that opportunity, I would think with this as well, to sort of vary what my interactions are like and learning experiences are like.

Dave McCool

Yes, absolutely. And I think the whole idea is agency. It's really putting you in an active role, not a passive role. Like you are that person. People are asking you questions, you need to react to them. You need to analyze that information and try to decide what the right course of action is as opposed to just consuming the information.

Michael Horn

All right, so as we wrap up here, what should we keep an eye on for in this space? If you put that Gates foundation funded researcher view back on. So not just Muzzy Lane, but just the world of simulations in general, where do you think this goes? Now with AI over the next five-ten years?.

Dave McCool

I think you touched on it earlier. I think more broadening the applicability of these across more areas. There are places like say HVac Training which really need a physical environment. They need 3D stuff. So pushing those AI boundaries to build those 3D assets for you. But I think simulation in my view really should be part of any real online learning experience. And I'll sort of make just a little plug for our AI principles because everyone has to have AI principles. For us it's three, it's to get learners thinking more, not less.

And you could see from that open response I had to think harder than if I had a multiple choice there. To increase academic and contegrity, don't decrease it. And I think again those active learning experiences are harder to cheat on. And third is we're really focused on making the online component of blended learning better, not replacing the in person instructional part of online learning of learning. So I think what we'll see is more complex, more interesting interactions, more subject applicability and just generally just more whether it's us or anyone else. Like there should be simulations as part of every learning experience. They are a great, cost effective way for you to practice, makes you better prepared for assessments, better prepared for in person apprenticeships and internships and all those things. So that's what we'd like to see.

Michael Horn

Beautiful. Hey Dave, thank you so much for joining us. We'll check in with you in a few years to see how that is bearing fruit. But really appreciate the work you all continue to do.

Dave McCool

Great. Thanks Michael.

The Future of Education is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Thank you for subscribing. Leave a comment or share this episode.