Staying Competitive in a Skills-Based World: The New Rules for Growth and Employability in the Age of AI

Ladies and gentlemen, the webinar will now begin. Over to you, Scott. All right, and thank you very much, Raven, and hello everyone. As mentioned, my name is Scott Ferguson, and I want to welcome you to today's webinar entitled Staying Competitive in a skills based world, the new rules for growth and employability in the age of AI. Now this event is brought to you today by HR Executive, and it is sponsored by Instructure. Now before we begin, I'd to go over some basic housekeeping items about the webinar console.

This event is completely interactive, and it features many customizable functions. Every window that you currently see from the slide window to the q and a panel can either be enlarged or collapsed. So if you'd like to change the look and feel of your console, please go ahead. If you have a question for our speaker today, please enter into the q and a widget on your console. That red widget which says ask a question is accessible on the bottom center of your screen where you see the question mark within the chat bubble.

Now we try to answer your questions either during the event or as part of the live q and a session toward the end. We will answer as many questions as possible, so we invite you to ask away. And if you do not get your question today, you may receive an email response after today's event. Now finally, Instructure has some additional materials that you can download and explore later. All you need to do is look for a blue widget that says related content.

And now for today's topic and the reason why you are here today, staying competitive in a skills based world, the new rules for growth and employability in the age of AI. Now as many of you already know, AI is rewriting which skills matter, and it's doing it faster faster than most workforce strategies can keep up. Now skills that mattered five years ago are losing ground fast, and many organizations are working to better align their workforce capabilities with their changing business needs. Now this isn't a story about reacting faster. It's about organizations building the judgment to grow their people on purpose and employees having a clear path to continue growing even as the ground shifts underneath them.

Now to learn more about this timely topic, we are joined today by Ryan Lufkin. He is vice president of global learning strategies at Instructure. And with that, Ryan, I'm gonna turn it over to you, and you can get us started. Great. Thanks, Scott.

I appreciate it. Thank you for joining us today. As Scott said, my name is Ryan Lufkin, VP of global learning strategy at Instructure. If you're not familiar with Instructure, we're the makers of Canvas learning management system, the most widely used learning management system in North America by k 12, higher ed, and increasingly in the corporate learning space. And so in my role, I actually get to travel all over the globe talking to learning organizations about their experiences.

And I've been in education technology for twenty six years. I've worked for Instructure for eight. And then I've been on my role with the learning strategy team for the last little over three years, really focused specifically on understanding the challenges faced by learning organizations, and really trying to focus in on how do we help predict the future to some extent. And as you can imagine in the age of AI, that's been probably a little more difficult than it would have been ten years ago. As we jump into the next role, the impactful eight.

So when when Melissa Lobel, who's our chief learning officer, and I started this team a little over three years ago, we were trying to understand the challenges faced at every stage of learning, from the first day of school to the last day of work. And we came up with what we call the impactful leap, really as a as a method for trying not to boil the ocean in what we were speaking about and and what we were honing in on. I'm gonna show you those really quickly. And and the goal really was to to create these pillars, understand these pillars that regardless of the type of learning organization, the stage of learning were applicable. And so we settled in on these eight, operational efficiency and effectiveness, right? Getting the most out of those systems that you use, right? We see, you know, in teaching and learning, a lot of churn in those technologies, and largely that has to do with failure to adopt, failure to get the data that you need, that kind of thing, lack of flexibility for your organization.

So how do we address those challenges? Right? Evidence based design, which really is evidence based everything. Data is the the key. Right? And so how do we get the data out of our systems to make sure that we are doing what's right for the learner, what's right for the educational program, and what's right for the institution at every level. Those are incredibly important. The assessment life cycle.

How do we understand that our learners are mastering the skills that we're trying to teach them? This is increasingly difficult in the age of AI, where assessment and the use of AI to circumvent assessment has become quite a challenge. And so how do we actually create more authentic assessment? More understanding of prior learning, understanding of the iterative process of learning, and make sure we're leading them down those paths. Lifelong learning. This is something we're going to talk a little bit more throughout here, but the days of going to school for twelve, fourteen, sixteen, twenty years, and then moving directly into a job that we stay in for thirty plus years are gone. Really, that was the previous generation.

And so how do we support that lifelong learning journey through every stage? And we'll talk a little bit more about that. Recognition of learning. This is traditionally the currency of learning has been the two year degree or the four year degree. And increasingly what we're seeing is people reskilling and upskilling with credential programs and specialized programs at the university level. And then a lot more of that corporate learning to upskill and reskill our employees, right? There's that adage of what happens if we train our employees and they leave? Well, what happens if we don't train them and they stay, right? And so how do we balance that approach? Education industry partnerships, so so closing that gap between the university level engagement and the the entry level jobs.

This is something again that's been heavily impacted by AI. Those those jobs that used to be, you know, the the entry level positions, those internships, a lot of them are being automated by AI. I know my first job working for an ad agency right out of college was writing radio bumpers for Coca Cola commercials for 10 Western states. That job is now being you know, instead of having five interns, that job is being done by AI much more rapidly. And so how do we create new pathways into the job space? And that relies on businesses and universities working together to make to bridge that gap.

And then universities and or employers in a lot of cases actually kind of picking up some of that training responsibility, and we see more and more, which is why many of you are here. Generative AI, obviously, you know, it's been almost four years, four years in November since OpenAI launched ChatGPT and launched the generative AI revolution. We're going talk a little bit more about where we are in that the evolution of AI, because it's incredibly important. But it really has been the most disruptive time in the history of teaching and learning across the board. And so how do we do we use those tools productively? How do we build AI literacy? Those are all major challenges.

And then something that's been really really focused on more recently is the science of learning, right? Instead of rote memorization, how do we understand how the brain learns, and how do we make sure that we create the cycle for learning and unlearning, right? I think it's funny because you know, there's there's so much talk about the the disruption of AI or AI being used instead of learning. Right? Yeah. We're we're losing the ability to learn in the same way we used to. Right? Cognitive offloading is a hot topic. I'm I'm now using AI instead of learning the task.

What's interesting is in four seventy BC, Socrates actually shared very similar concerns. He saw his learners moving from rote memorization to writing, and he was very concerned about that being disruptive to their ability to actually gain and retain knowledge. Right? So this is not a new conversation. This is a this is a twenty five year old 100 year old conversation we continue to have, and at every stage we feel as though humans are dumber. In reality, we're we're actually much more more intelligent.

We're we're much more productive when we work with these tools. And so we're gonna talk a little bit more about that today. This is a slide I actually love, and even before COVID, I was speaking to exactly this kind of transformation, right? The global workforce, if the global workforce were 100 people, 41 would not need upskilling by 2030, right? They're going stay roughly the same. their current role, right? That's a lot of on the job training that reskilling position. and redeployed, right? So they're moving to different jobs.

Upscaling or rescaling would be a term I would use for this, that rescaling them for different positions in the organization. And 11 would be unlikely to receive the necessary upscaling, and we see this a lot, right? Where companies don't have the training programs in place to help their existing employees change into new roles, right? It costs a lot more to train a new employee off the street than it does to train an existing employee. And so how do we make sure that we have those programs in place? And then the skill disruption. This is something that was hotly debated within the education space prior to COVID, and even more so now with AI. But that disruption of the skills that we learned, as we all look back on our learning journey, a lot of the skills that we learned were either taught ineffectually or are just no longer applicable.

So 39% of workers' core skills will change by 2030. There were a lot of people in the academic world that didn't think this was the case. Right? They were talking about human skills versus technical skills and trying to get very distinctive on on what those skill differences were. That conversation is still going on, but what we understand is the skills we use in the workplace are different and will continue to change because we're in the middle of of massive disruption like we've never seen before. We're right in the the center of the storm, and it's not likely to change in their future, but we all get to shape that.

It's that it's one of the reasons I get so excited about, having these conversations is we actually get to shape what the future looks like, and we all play a role on that. So it's pretty exciting. One of the things that's been interesting is, and we'll talk a little bit more about this in a minute, is AI is already entering the workplace. Right? And what we've seen is early productivity gap. Right? This idea that that we're not exactly sure how to apply these tools.

A lot of claims are made, bold claims about jobs being replaced by AI, things like that. And actually what we've seen is an actual dip in productivity because of that, right? How do we use these tools? How do we use them in innovative ways, right? There was a there was a really interesting speaker I saw recently who said, you know, when when factories were originally built, they were running on steam, a steam plant. And when electricity came along, the industries that failed really were the ones that just simply replaced the steam engine with an electrical output. And the ones that succeeded were the ones that actually looked at how electricity could be used to evolve their processes themselves, and actually did major innovation. And that's what we're seeing today, is this this productivity dip in in anticipation of what will be productivity enhancements.

And so that's why it's heartening when we experiment with AI. One of the things I always say is we need to build in failure into our innovation. We need to build room for failure, so as we experiment with these tools, as we try new things, we have the opportunity to learn from what didn't work and build off of that. Something from a business standpoint, I think, is incredibly important. And most of us understand this piece, but the the capital expenditure just in $725,000,000,000.

This is more than the 3 Gorges Dam, the entire US highway system, the Louisiana Purchase, right, the Apollo space program. More money is being spent to build AI infrastructure in this year than in all of those projects. That's pretty incredible. And so AI is not a fad. It's not going anywhere.

We will continue to see its growth. And in reality, think that's incredibly important to acknowledge, not just because, you know, I happen to live in Utah. There's a major data center debate because we're in the middle of a desert. Data centers consume a lot of electricity. So there are concerns.

There are valid concerns about this, making sure we we build the infrastructure in the right places, making sure that this is a shared expenditure, not necessarily just being put on taxpayers, things like that, but it is happening. And so how do we make sure we evolve with that process? There are a lot of very grandiose claims being made about AI. And I was actually in Australia speaking at a conference when this Microsoft AI's chief came out and said, all white collar jobs will be automated by you know, in the next eighteen months. And I looked out of a room of about 300 people and said, that's your jobs. Do you think your job's going be gone in eighteen months? And they all laughed.

The reality of this is this is what we call AI washing, right? These these grandiose headlines of workers being wiped out and replaced by AI, they're just not realistic to how AI works. They're great for boosting valuations of AI companies and things like that, but the reality is, know, these claims aren't really made based off of the reality of pace of AI adoption. And so we see these headlines, we recognize that those cause anxiety in in our employees, in our future employees. Right? There was a there was a rash of university commencement speakers who were booed when they brought up AI this last this last graduation season. And the reason for that is they see headlines like this, and they think there's not going to be a job for me in the future.

I will tell you there absolutely is a job for you in the future, especially if you understand how to use AI to make yourself more efficient, but we've got to help them see beyond these these disruptive headlines. It's incredibly important. That said, this is by Anthropic's own data, this is where we see AI already impacting jobs, right? This doesn't mean eliminating jobs. This is an important conversation, but impacting jobs in the workplace. And and blue the blue outline is actually what they the areas where they anticipate AI having an impact on jobs.

The red is where we're already seeing that impact, right, based on data from last year and earlier this year. And so this is incredibly important to understand. Like this is not every job is not going to be replaced by AI. That's unrealistic. But we can actually start to see where these tools are being adopted and for what.

And when we dig a little deeper, we can actually understand the ways that these tools are being used to make people more productive. And there's there's there's a full report around this. Anthropics done a great job of generating good data and sharing that data in insightful ways. If you're not already following some of their work, it's worth worth following. What I love about this, and this is the Boston Consulting Group's report about AI driving high level skills task automation.

Right? And this is the way we need to start looking at this, is it's tasks that are being automated, is not roles being replaced. And this is incredibly important. I think we see a lot of those headlines, as we mentioned, talking about AI replacing jobs. We see a number of headlines about companies laying off employees, and saying that's because of AI efficiencies. Generally, it's not.

Again, another form of AI washing. Those companies haven't achieved the level of automation with those AI tools yet to actually achieve that. It's an unfortunate kind of of pinning that on AI because it it it's just not accurate when you actually dig into the data. And and we need to be looking at these these these task automation. One of my favorite aspects of this conversation is this this attempt to try to pin these irreplaceable human skills.

You know, we immediately after AI, they're like, well, AI isn't isn't capable of critical thinking. Actually, it can. AI isn't, you know, capable of creativity. Meh. You know, to some extent, it actually is.

There's some seen some pretty funny music videos that it wrote the music for and and lyrics for. It can. Right? So dismissing those things out of hand is the wrong way to approach AI. Right? I often talk about AI, you know, when we were first talking about AI tutors and things like that, one of the one of the best versions I heard was AI is like a lazy graduate assistant, right, for a college professor. It does just enough work without really kind of being understanding where it needs to go on its own.

It needs to be directed. It needs very specific rules and and guidance. With AgenTik AI, it is much more capable of following those tests, but it's still running down a task that is being given by a human ultimately. And you can argue with me that that, you know, if you start daisy chaining AI agents together, but ultimately, there's a human at the top of that funnel giving those job giving those tasks and and trying to outline what's possible. So that's incredibly important.

And so, as we talk about those those irreplaceable human skills, realize we shouldn't actually be talking about any human skills being irreplaceable. It's the human directive and the human human guidance that's incredibly important this process. That said, there are really important skills at play. Right? There was a there was a professor at California State School. He was on a on a panel with me, and he he actually outlined a really interesting shift in in how they were using AI in the class.

Right? He had a product design class, and he said, you know, it used to take us 90% of the class to define a product that you want to explore, do some basic packaging design, do some competitive analysis, do general market analysis, come up with kind of a general marketing plan, and then they would spend the last 10% of that course presenting to other students. And what he found was with AI, most of that work can be done within 10 to 20% of that course. So now he actually makes those students do some industry research, find industry experts, reach out and schedule a call with those experts, present the the their their project to those experts, and then get their feedback. And I have students. I have a college aged student, child, college aged child, and a high school aged child.

And what's so interesting is, you know, if I call one of them, they will text me back and say, what do you want? Right? They won't answer their phone. It's scary for them to answer a call from somebody they don't know, right? And so leaning back into those more human skills really is the value of that course, right? They're leveraging AI in much the same way we would in the business process, but they're leaning to those more human skills of making connections, presenting, communication, network building, those kinds of things. Those are incredibly important. And so as we start thinking through how to use AI within our own programs, very valuable lessons to be learned there. So not every skill is at equal risk.

There are skills that are automatable, absolutely. They can be end to end, right, like the basic radio blurb comment I made earlier, that's a great one. Right? Basic writing, the summary of of materials, basic research stuff, those are all actually really easy to automate. Right? The augmented piece where AI speeds it up, and and this is where we see, I the vast majority of this. We're going to show you a really interesting chart next.

But then those human led, where trust me, accountability don't delegate. Right? And this is important because we still see I actually had fellow executive of my company send me a list, and he said, hey, did some research on all of the AI tools being used by universities across The United States. You might want to use that. It's great. Right? I was like, oh, wow.

That's a that's a really incredible list. And I looked at the first list. Some of the University of Michigan, Maisie. Oh, yeah. I'm like I know there too, whatever.

I looked at the next one, I was like, that's not University of Notre Dame. It was like leprechaun something. I was like, that's not that's not real. And I do I might have did I miss that? It was at some development I hadn't seen. Right? And what I found was, you know, he had populated that first list of the first line of the set of the of the chart and said, give me all this information for all of the the AI tools being used by universities across The United States.

And the tool, in an attempt to give him what he wanted, made up everything, made up 99 other examples, completely from scratch, made up URLs, made up names. Some of them were very catchy names, right? But it wasn't real, and that's one of those things where we always have to be critical of the outputs of these tools. We are the subject matter experts. We remain the subject matter experts. We've got to be able to to provide that that level of input.

What's really interesting is this chart from Boston Consulting Group, and I actually love the way they broke this out. You'll see the from the bottom, from the left to the right is the level of disruption we'll see. And again, disruption does not mean loss of jobs, and I'm going walk through each of these a little bit because I think it's incredibly important to to dive into these. Right there in the middle, you see amplified roles, right? And they only put this at about 5%. I would actually put it higher, but these are also like software engineer, right? Where they're actually maybe no longer writing the code, but they're directing multiple agents to write their code, right? Fundamentally, that's part of their role, and they are just using these tools to amplify their outputs in really powerful ways.

I know a number of programmers that fall into this category. They actually Boston Consulting Group says, you know, many lawyers, particularly those in advisory and judgment intensive areas fall into this category as well, right? This is kind of where I think we spend a lot of our time. Those rebalanced roles, like this is one of those where marketers fall into this a lot, right? So instead of being broken up by channel, which a lot of them are today, they'll expand their roles to be more of those omnichannel specialists, Right? And they'll they'll run end to end campaigns with all the analytics and all of the elements that they used to have to rely on decentralized role. I came from the marketing world. I understand this really well.

Suddenly, they're much more self contained. They can do some of the creative themselves, whether it's whether it's visual or writing. They can do that that planning. They can get data out of the systems that they're working with and aggregate those new reports. Right? All those things used to be much more difficult.

Now you've got that omnichannel approach, right? That's an incredible role, a critical shifting of those roles. What is the divergent role, right? And the example they use is insurance agents, which I thought was fascinating, where so many of the actually mundane tasks are automated. You know, they've got lead qualification, engagement with claims, all of those types of things. And that saves them a lot of time, but there's still a level of critical thinking, a level of of balancing whether or not they're the right clients that that has to be in play. Right? But they can actually use their time now to support audiences that they may not have before when they had to spend more time.

They can expand their audience base. That divergent role is spending your time on different things in in pretty remarkable ways. I think the the the substituted role is probably the one of the biggest layers of disruption. So they they call out certain financial analyst roles, right, where the volume of the reporting and the analysis they're doing, a lot of that just can be done by AI, right? These roles are still rely on the internal decisions based off of that data, but that reporting and the the research they were doing before, that's done in seconds, much faster than what they would have otherwise. Right? And so they they can, move these productivity gains into more productive areas.

They can expand the number of analysis they're doing. They can make themselves more productive. But ultimately that core role that they were doing is now automated. Right? And that's one of the most disrupted areas. We see this in a lot of potentially different roles, but not as many as you think.

And I think a lot of these are more divergent or rebalanced than being substituted, right? The enabled roles, I love this area. The enabled roles I think a lot of us find ourselves in, where, you know, and they call out clinical assistants and lab technicians, because I think some of the diagnostic work, this is another input, right? This is additional input, it makes it easier to get information, but their role fundamentally has not changed, right? They they're using AI in real time to to give them more efficient information, but they're still doing essentially the same job. I think a lot of us probably sympathize with this role or or or identify ourselves in this category. And then the limited exposure roles. Right? They they talk about physicians and teachers.

Right? Where ultimately, they may be using AI tools to make their job more smooth as the administrative side of their job more seamless. Right? Those administrative tasks can be offloaded, note taking, grading, some of those things can be offloaded, and they can really focus on the human connection between doctor and patient, student and teacher. That is the connection. That is the core of their job. They get to lean into that and do it do it more effectively, which I think is actually the way that we want to see the world, right? That's the way we want these tools to work, and can be very powerful.

This Boston Consulting Group report, I actually link it in the slides. Think Tammy in the question Q and A is asking if we're going share out the link to rewatch. We will share out the link, and we will also share out the slides with links to these resources. So I think there's a lot of great resources from groups like Digital Education Council, Boston Consulting Group, and a lot of others, UNESCO, and some of the work they've been doing that are incredibly helpful to dig into this. So one of the aspects too is there's I'm going to identify with a lot of you.

You are probably getting directives from your executive teams to increase use of AI. Not a lot of guidance on what that looks like, right? And that's challenging. We're seeing that across the board from universities to private sector to public sector, right? There's this idea that other people are finding efficiencies we have to too. In a lot of ways, I want to make sure that we're not being reactive, right? We saw this during COVID. We were reactive to online learning.

We threw our in person courses into online learning. And then we wondered why the engagement wasn't there or the quality of education wasn't there like it was before, right? Because they were not intentionally designed. We're at risk of doing the same thing now with AI if we're not careful. And so I want to I to I want to draw a clear distinction between that reactive side where you're fixing gaps that, you know, that are already a problem, one off training, HR working in a vacuum, the more intentionally designed approach, which, you know, a standing plan for you know, before the gap is created, looking at what the future looks like and creating a coherent plan. Anchor to real business goals, this is really really important.

What problem are you fixing? What goal are you trying to achieve? How does AI help you get there as opposed to applying it for the sake of applying it. Right? And then pick the tools after the need is clear. And this is what's interesting is we see a lot of enterprise wide decisions being made around the big partners. Right? You're probably a Microsoft company or a Google company, right? You use Microsoft Outlook, you use Google, Gmail, that kind of thing. And in all likelihood, you probably signed your company probably signed an enterprise wide deal with Gemini or Copilot.

In some cases, you might be using OpenAI or, you know, other tools that are out there or Anthropic. And in reality, what we're starting to see is while those are great enterprise wide tools that help increase AI literacy and and a general adoption. Those are great. They're not the best tools for specific roles necessarily. We're seeing this across education, whether it's higher education, whether it's corporate space.

There are tools that are often better at those specific tasks. Those large language models are trained on specific content. They're they're better fit for task. And so now we we move beyond that kind of top layer decision making. Now we've got to start making decisions around that those next layers of tools.

We've got to make room for those tools as well. And that's really important. Picking those tools with a clear understanding of what we're trying to achieve is incredibly important. And then what the organization owns, right? Building governance about the, you know, before you scale not after. Right? We've seen a lot of this where, you know, we signed an enterprise wide deal, we rolled it out to all employees, and then holy cow, we got hit with the bill.

Right? These tools aren't aren't free, you know, unless you're using an open source tool, and even then you're paying to host that and all that kind of thing. The the tokens involved are not cheap, and so having an understanding of that, you know, trying a beta group and really understanding what their uses cases are, Because we've seen even within Instructure, we've seen individual groups blow through their token allotments really rapidly because they're using it in innovative ways that we didn't anticipate. So how do you how do you balance that? How do you understand that before doing that enterprise wide rollout? Invest in training. You know, just rolling out these tools to people. People will default to using using them as a a smarter browser.

Right? They're they're a smarter search engine, and that is a good use. It's basic use for these tools. It's not the agentic phase. It's not using these to accomplish tasks. It's not replacing workflows.

It really isn't driving more efficiency. You might get a little bit more accurate answer than than Google, but you're not fundamentally driving efficiency within the company. Right? And then Inger's heels work to a real business goal. We're gonna iterate this multiple times. Right? Because solving real problems is what these tools are good at, not just applying them to, you know, being the electrical version of the steam engine.

That's not the answer. Right? no formal AI governance. They simply don't know what the rules are. And I will tell you, you know, I'm gonna I'm gonna use my kids as a as an analogy here, but my daughter who is in college, she's a senior. She was a COVID kid.

Right? She's a rule follower type a, and she didn't start using AI in any way in her coursework at all until this last year because she was so scared that would she would be flagged for cheating or she would, you know, she would be she would be viewed as using it incorrectly. And she wasn't being given the guidance, and so she didn't understand. And and so she just defaulted to not using it. Finally, she had some professors who were saying, hey, we're going to build this into the course. We're actually using this as part of the process.

Here's how you use it ethically and effectively, and that changed her mindset towards it. My son, who's a sophomore in college or in high school, his approach was, if I can use Grammarly, why can't I use ChatGPT? Right? He just sees it as another tool. There was no guidance given. Most of his educators at the time were just saying, this is cheating. Don't use it.

He's like, how's it cheating? I can use I use autocorrect. I use all these different tools I don't understand. And so that level of guidance and training is important. Don't assume a level of knowledge that you have applies to everyone else. Right? That's incredibly important.

AI related training yet. That's crazy. Right? Almost half don't offer any kind of training. And this isn't just roll out the tool and go Google Yahoo, is what are your expectations as an organization around how these tools are being used? Incredibly important. And again, this goes back to that lifelong learning element we were talking about before.

And I love Alvin Toffler's quote, the literary of the twenty first century will not be those who cannot read and write, but those who cannot learn, unlearn, and relearn. This is incredibly important, the ability to unlearn. And you see that, you know, I showed the in productivity, right? That's that unlearn mode right there, right? Right there at the beginning of this. And how do we unlearn the ways we were doing things before and rethink these new processes? And this is something that we're being asked to do throughout our life cycle, and it's not going to stop until we're done working, right? This is the pace of change around these technologies. While for, I will say for AI, it's plateaued somewhat, the early gains we saw in the first really three years have slowed, right, as these as these models models were being trained on millions and then billions and trillions of of parameters.

And now we said, we we kind of we've trained them on all of human knowledge. How do we then actually back that down and find tools that are more fit for purpose, consume less energy, a better level of processing that makes more sense for these tasks. Right? And that's kind of where we live. But that'll continue to change as we move forward, as we as we move into we'll talk about the agentic phase versus the other phases because it's really important. And this isn't going to stop.

So your roles are incredibly important at enabling your employees to your learners to create that ability to learn and relearn, right? Super important. Designing growth on purpose. Right? So again, this is this is intentional versus versus, you know, reactive. Right? From an organizational standpoint, build the standing plan, pilot in one team first. I can't I can't call the importance of this enough because you you learn so much in those pilot groups when you get feedback from them, and again, ask them for feedback.

And then tie learning to real roles. Right? Not everybody's going use these tools the same. Right? Your your marketing team uses them differently than your development, your programmers. Right? Different than your business analysts, business, than your go to market or sales teams, right? These are all going be used in different. Understand those different use cases, and understand that applying a single tool may not be the best option, right? The skills that they're going to learn are going to differ around AI as anything else.

Name the skill you're building next, right? This is one of those things where people understand where they're going. They're much more likely to engage in the journey. It it it is scientifically proven that they understand. Ask for feedback before you need it. Right? This is something that too often we assume that things are going well, too often we assume that people understand.

Always engage in that feedback loop. That is incredibly important, and listen to the feedback. Too often, we all have the canaries in the coal mine who might be more vocal than their neighbors. Sometimes we tend to write off some of that feedback, don't. Understand that they may represent the feedback from larger groups.

And then treat unlearning as part of the job. Like I said, incredibly important to create that unlearning, because it's not just about applying these tools to an existing task, it's about rethinking the methodologies. That's where we find real innovation within organizations, and then we've got to create room for that. Right? And and again, the ability to fail, the ability to try and fail is incredibly important there as well. This is what I like to to call out a lot, because I think when you actually understand that this is we're not in the eye of the storm, we haven't gotten through the AI storm.

We're in the the next phase of the AI storm. Right? So perception AI is speech recognition. Right? This is this really was machine learning and the the early phases of AI that could do simple chatbots mentality, that could do basic research, things like that. When people talk about AI, there's AI before there's AI after. Right? And AI after is generative AI.

It is fundamentally a different thing, and that's where large language models took over that thinking. And large language models are a bit of a black box, right? There was a great quote from an AI executive that said early on that AI, large language models could be three raccoons in a trench coat. We don't know, right? And the idea was really getting to the lack of understanding of what was happening in there. Anthropic and some of the other large language model providers have actually done a bunch of research trying to understand how large language models actually make their decisions and create their outputs. And so your generative AI mode was really the fundamental change in writing and creativity that we've seen over the last little while.

The phase that we're in now really is the agentic AI phase. And the like, this is where agents are conducting work on our behalf. Right? And this is much more empowering. There are potential risks, right? There's an open source Genetic AI tool called Claw that was being used by a couple there was a there was a user that said, hey, go make me a dinner reservation. And the tool actually went out, tried to make a reservation online, and the restaurant's reservation tool was broken.

So it gave itself a voice. It called the restaurant, and it made a reservation for them for dinner. And people then proceeded to freak out because it did something unexpected. Right? It in attempting to accomplish the task it was given, it circumvented the the traditional route to do that. Right? In the same way that AI made up those results trying to give my coworker what he wanted, It wasn't nefarious.

It wasn't an attempt to to, you know, deceive him at all. It was simply trying to accomplish that task. And when it couldn't do it with the with either the time or the the access that it had, it made it up. Right? And it didn't understand the difference necessarily between making it up and and actually finding real results. And some of that's in how you prompt the tool, but a lot of it is tools have gotten smarter and they're less likely to do that, but they still do.

And we've got to make sure that we anticipate that there may be the unanticipated. I know that's kind of a quandary, but sometimes these tools do the unexpected in their results. We've got to understand that. We've got to learn from that. We've got to make sure we build guardrails on privacy, security, accessibility, those types of things to make sure that that these tools don't go off the rail do things that we don't intend.

But the reason it's important to embrace the, you know, this this approach, understand, you know, AI literacy, and I and I always tell people, it is okay to be an AI detractor. There are plenty of of concerns, whether they're environmental concerns, privacy concerns, copyright concerns. There are those concerns. But in order to be part of the conversation, you have to have basic AI literacy so that you understand what to be concerned about, and don't assign that that concern to the wrong categories. Right? But moving forward, we're going to see more of what we call the physical AI aspect.

We're already seeing those. I wrote a Waymo when I was down in Tempe, Arizona a couple of months ago. Other than the fact that it got stuck behind a cat walking down the middle of the street and wouldn't honk its horn, it was a it was a seamless experience. It had great music. I sat in the back and relaxed as much as you can do when riding in a robot car.

But that's that's the first phase. Right? This is something we've been working on for a while. We're going see more and more of that. We within our lifetime see assistance robots, quadrupedal humanoid assistance robots being deployed in our communities, right? Where China has actually got a little bit of a lead on us there. They're looking at it as a means to address the little kings generation, a generation where, you know, large number of of unmarried men didn't create families that would support the family, multi generational family structure that they've had.

They're looking at robots as a way to come in and help the elderly in a situation where they don't have children or grandchildren to do that, right? They may be a little ahead of us. It's it's happening. Sounds like science fiction, but it's becoming real, and we need to be able to understand how that works. So the value of education shifts from delivering knowledge to developing capability, right? And this is incredibly important, because I think across the board, idea that going back to my first thought on Socrates, Socrates believed that by by passing his knowledge down to others and having them memorize it verbally and and through that method was the only way to really create true intelligence. Right? He truly thought people would be dumber for having written things down.

And now we're having the conversation that we need to write things down, not type them, because we're gonna be dumber if we if we type them instead of writing them down. It's a it's kind of a it's kind of an ongoing, you know and and you could say that humans are are less intelligent, but you'd be wrong. Right? Humans are massively more intelligent than they were two thousand five hundred years ago. And so clearly, we're we're doing something right through this evolution as we work through these tools. I wanted to to talk a little bit through designing for the AI learner, because it's I know I'm going a little bit longer than I wanted to to get here, but we'll go through these.

Designing for the learner today is different, because they are doing assessments differently, they have access to AI at all times. Know, we have we have employers, universities who are doing lockdown browser when doing assessments and things like that. Well, pull out your phone. Right? There's a there's a tool called Goth AI that is literally you can point it to the screen, and it will give you an answer to the quiz that's being presented on screen. Things like that, right? There was a I saw there was an employer who was sending out a mirror adviser to to potential employees, so that when they were doing their interview, they couldn't hide, you know, a phone screen or another screen off of the the tool and use that to answer questions.

That's alarming. Right? And and we should we should have to be understanding that these these potential employees are using AI. They're going to. You're going to want them to use AI to to become smarter. But how do we actually understand what their true like skills are in the process? And so develop and learn for that is really important.

And these are the three agency, judgment, and adaptability that we're going to walk through in a little bit more detail. But these regardless of the level of your learner are incredibly important, right? Whether you're K through 12, whether you're higher ed, whether you're corporate learner, but building an agency, right? So not just compliance based assessment with project based not just compliance based assessment, but more project based, right? This is something that there's a lot of data for within the higher education space, where project based learners are much more likely to embed a skill, because they actually understand how that skill as opposed to being ethereal or, you know, arbitrary. Certainly, they understand how it's applied in the workplace, and that makes it click in their head, right? Making sure that we're teaching that in the same way that project based piece is incredibly helpful, right? Give learners choice and pathways tools and demonstration formats, Not every learner learns the same. We see this a lot more there's a lot more focus on this in in schools at every level at this point, but different learners learn differently. They need accommodations for that.

How do give them some level of choice? Right? There's great tools like Notebook LN that can take all of the learning materials within a course. They can ingest them, and then the student can actually say, I want flashcards. I want a podcast. I want to listen to this as a podcast. They've had read through the podcast, I want to ask questions.

You can do that too. Right? So there's great tools out there that support this kind of flexible personalized learning experience. Make metacognition visible, require reflection on what we learned, not just what was produced. There's a there's a great tool out of University of Sydney. They they have AI startup that was built in house called Cogniti, and they've got a little tool that happens to live within Canvas.

They can build simulations directly within Canvas itself. These are these are they're vibe coded. Right? And vibe coding sounds hard. Vibe coding is literally describing to an AI tool what you wanna build, and then how you want that output, and then it builds it for you. Right? And then you can iterate on it.

It's vibe coding is not as hard as it sounds, nor is technical. And that tool actually there's a there's a great tool that you drop in and you you write code, and it's giving you real time feedback. And then the way they set it up, your instructions are on the left, your attempts are are in the middle, above that is what what you're getting wrong, and so it's giving you real time feedback, and it'll let you iterate on that process. Right? An incredibly valuable tool to be able to iterate in real time with that feedback, a great approach. And then teach AI uses a learner skill.

When to use it, when not to, how to direct it. This is incredibly important for every organization. Basic AI literacy, most of us are using these AI tools at this point. Something like 85% of all employer employees are using AI at some level within their personal life. Right? We need to make sure that they understand how to apply it within the workforce, and what your expectations as an organization are.

They don't assume they know. And and again, reiterate it every time, because you might do that in a in in the a company all hands or something like that. That's great. The employee that started two days after the company all hands isn't necessarily gonna hear that. Repeat it multiple times.

Right? The old adage that you need to hear three things. Alright. Here's something three times before you learn it. It applies today. Develop judgment.

Right? This is something that's incredibly important for the future. So as these tools evolve and become more powerful, they need that critical thinking by the user to be successful. So build assignments where learners must critique, correct, or extend AI generated work. We're seeing a lot of this in higher education, right, where they're they're flipping the classroom, flipping the assignments, and having AI generate the work, and having AI having students be the critique, right, as ways to to actually measure knowledge and not be able to to circumvent that with AI. It's an incredibly interesting way to do it.

Anchor judgment and disciplinary expertise, learners can't evaluate what they don't understand. Right? They got to understand how the basics AI literacy works, that kind of approach. Use case based learning ambiguity and trade off decisions as the core unit for practice. Right? This is interesting because it's actually goes to basic problem solving, which is fundamentally one of those things that I know I said no skills were human skills were replaceable, but it is one of those differentiators. When you look at Bloom's taxonomy revisited, there's work that's been done by a number of universities.

Oregon State University did some great work with that. But there are those those basic problem solving approach. Right? You can use AI as an assistance to that, but the actual human ability to solve problems, to understand situations is really unparalleled. And then assess reasoning not just outputs. Ask learners to defend the choice, not just deliver the answer.

Right? This is as old as time. Show your work. Right? This is something that when you show your work, when you have to when you have to catalog that, you're much less likely to turn to these tools. But you're also even if you use these tools, you're able to understand how they use them, why they use them the way they did, and learn from that. That's incredibly important as well.

Cultivate adaptability. Right? Embed transfer skills, transfer tasks, take a concept, learn in one context, apply an unfamiliar one. Right? This is that human flexibility, that ability to to roll with with changes in the organization, with the market. This is just incredibly valuable skills for your company as a whole, and you can learn to actually apply that in the in the in the process of developing your training. Design for unlearning, again, that that true thread you're gonna hear multiple times.

Show learners how to recognize when a prior method has stopped working, and then be innovative with how they approach that problem, right? That's really important. And if you look through, you know, back through your own history, you will think of times when you were doing something a certain way, it was no longer effective. You thought through the process of doing it differently, and that's where we find a lot of our key successes, right? How do we help learners understand that that is okay to do? I think there's a lot that's been said about younger learners and their ability to adapt to change. A lot of a lot of learners struggle with that. Right? A lot of new employees can can go from from a to z if they're given every step along the way.

They may or may not be able to go from a to c, and that's not to paint them all with one brush. Right? Because learners are very different. Employees are very different. But making sure that they understand that that ability to unlearn is is core to who they are is is important. Build feedback loops to continue throughout the career.

This this goes you all know that. That's something that those feedback loops are invaluable for helping employees, understand their pathway, understand what they're doing right, what they're doing wrong, how to improve. That's incredibly important. And then making learning agility itself measurable. How do we how do we apply measurement to these different things? And there's lots of different softwares.

You're all trying to apply this at some point, but making sure that this is all being captured for for the improvement of your organization. So the shift isn't a total rollout. It's an institutional rethink. Right? This is what's important is, you know, I I come back to the analogy again and again, and and maybe you need find a a better one for for different groups, but that ability to just apply techno to apply electricity to a system that was run on steam is not the same as true innovation, and where you're figuring out how to actually fundamentally modify with electricity your entire process, right? The same thing applies. And so I'm going to jump through these really quick.

So I want to make sure if we have time for questions, please drop those in the q and a at this point. I think it's incredibly important that we keep this as a dialogue. We learn so much from our peers, from what they're doing, from from experts in the field. There's so much noise when it comes to AI just across the board. We find ourselves in a position where we need to make sure that we cut through the noise, find out what really matters, what's applicable to us, our organizations, we apply those.

So this isn't a tool a tool rollout since there's no rethink. We've talked about that a couple of times, but restrictions first AI policy to guidance and integration. We went from a lot of companies, a lot of schools, a lot of training moving directly to Ban AI, Block AI. Right? That was kind of ridiculous. It's kind of when we look back and we're like, wow, we really blocked AI across the board? That seems silly.

Right? Now it's this guidance and integration. How do we actually be much more directive with that? Compliance based assessment versus capability based evidence. Right? Collect that information. Move beyond just the compliance check boxes that that is is been a movement for a while, not new. The diploma or degree as the endpoint.

You know, I'm a firm believer that the the degree holds high ROI value, that it's a great way to measure a student's readiness for the workplace, but there's a lot of of additional, you know, stackable credentials that now as we move towards this lifelong learning journey, we need to recognize those. And how do we how do we build recognition of those into the hiring process? How do we support our our employees and potentially going out and seeking those from third parties, from universities, things like that. So that's the the one stage along the lifelong path. And then educators as content authority. Educators as judgment coaches.

Right? I think that the we move beyond that content authority, that subject matter expert role. All that information is out there and and and explorable. I tend to think of this when I speak to university audiences. I tend to think I tend to speak more of of educators as guides. Right? Guides through the knowledge.

You're there to help them along that journey. You're no longer just the sage on a stage, which is really important. So the ground will keep moving. We need to build the foundation to move with it. And this is what's hard in our role, is really understanding how do we make sure that we are prepared for tomorrow.

And and this isn't this a fad that will pass. I often throw up, you know, there's a Homer Simpson quote, you know, the Internet, is that thing still around, right? And there were headlines at the time. I have slides in another presentation that show that people very much thought the Internet was a fad and was going to go away. Think there are people that still think AI is is in that same vein. It's not.

And so how do we make sure that we're leveraging it for success? So that is all I have for today. If you want to connect with me, here's my information on LinkedIn. I host a podcast called Educast 3,000 with my boss and friend Melissa Lobel, who's our Chief Learning Officer. And then there's a lot of resources out on our blog. I write a lot about AI and learning as well as some other incredible folks on my team.

You can find that out there as well. And then I think throwing this back to Scott, I think Scott's going introduce some additional resources that are available as well. Alright. Ryan, thank you very much. That was great.

Appreciate you being here today and going all through that with us. As Ryan mentioned, that's kind of all the time we have for today. I wanna thank everybody in our audience for joining us. We hope all the information was useful to you. It was to me.

Again, thanks to Ryan for his insights today and, of course, the Instructure for their support of today's event. As a reminder, Instructure has some additional materials that you can download and explore as Ryan mentioned. And if you missed any part of the presentation, it'll be archived on the HR executive site so you can view it again, or you can refer someone else to it. And while you're on that page, we want you to look at other upcoming and on demand presentations, that we have going into now this last part of 2026. Once again, thank you for all your time today, and, have a great rest of your afternoon. Thanks, everyone.