Episode Transcript
[00:00:00] Good morning everyone. Today I want to highlight the really significant research we're launching this week on corporate learning. I'm not going to give you the whole details on the podcast. You can read about it in Galileo, you can read about it in the article I just wrote.
[00:00:15] But this is a massive, massive thing. Corporate training or knowledge management is at least a $400 billion market in training alone, roughly $1,400 per employee per year, distributed in various different tranches, depending on the level of the person.
[00:00:36] And our new research, which we've been doing for two years, finds that 74% of companies say they are not effectively keeping up with the skills needed in their workplace. In other words, one quarter of of the companies we surveyed, and this was more than 700 organizations, mostly larger companies, do not believe their training function or process or knowledge distribution approach is working. Now, there's good reasons for that. One reason is the pace of change is so high that nobody ever feels like they're keeping up.
[00:01:12] The second is AI, which everyone feels they're behind on because it's changing every day.
[00:01:18] But the third is that the operating model and the tools and technologies we used for learning are decades old.
[00:01:25] And we hypothesized and wrote about this in our Revolution of L and D paper three or four months ago, and now we've proven it with our definitive guide. As many of you know, we do these studies regularly. This is the fifth or sixth time I've done a big study of corporate training. And this involves hundreds of interviews, discussions with CHROs, HR leaders, learning and development professionals, and other business people, vendors, and a massive survey and analysis of the data. And what we've discovered is that we believe AI is not only revolutionizing the current process of training people, which I'll discuss in a minute, but it is bringing us to a new era, the era of dynamic enablement. Enablement of people, not just training. And we use the word training and learning a lot in companies because it's a well understood concept.
[00:02:22] But the problem with the word learning is it takes you back to the pedagogy of school where you're sitting down being spoon fed some sort of information or course.
[00:02:35] Yes, that's useful in some situations, especially when somebody's brand new to something, but it is not that useful on a day to day basis.
[00:02:46] I'm late in my 60s here and I haven't taken a course as a structured course other than compliance for probably 20 years. But I am learning things every single day from YouTube, from other people, from peers, from conferences, from My own reading, from my own podcasts. I listen to everything.
[00:03:09] And that is what happens at work, day to day, month to month, as things change, as the business changes, as the technology changes, as the customer market changes, as the regulations change, There's a continuous need to educate and inform people.
[00:03:26] And AI is exceptionally good at this. The proof point is very simple. 900 million people a week use ChatGPT. That number is probably higher now. And the data from OpenAI shows that more than 60% of them are doing this to learn something.
[00:03:44] That volume of consumption is orders of magnitude higher than all of the online training companies that have ever been created.
[00:03:53] And this is only two or three years old. The programs and the technology is getting better.
[00:03:58] And AI native learning, and I'm talking about AI native learning, not using AI to build traditional courses is exceptionally powerful.
[00:04:07] We know this because we do this in Galileo. You can use AI to build a course from a document. You can use AI to complement a course or change a course from a video or an audio. Ask the AI to build a five minute course, a 30 minute course, a one hour course. You can prompt the AI to add more interactivities or a case study. You can prompt the AI to turn the course into a podcast. You can use the super Tutor to ask questions. And every course or every body of knowledge that's added to an AI system informs everything else.
[00:04:44] So even though you may be reading or listening to one topic, you can ask the system about any topic. And they're all connected together because of the magic of the embedding technology in AI. So this is a completely different paradigm than traditional training. Now, I've been studying learning, as most of you know, for, you know, almost 30 years here. And one of the things that's always occurred to me in this space is the amount of, maybe the word is dogma or traditional thinking that's been applied.
[00:05:18] There's thousands of books and lots and lots of smart people that have come up with approaches like blended learning, spaced learning, experiential learning on the job learning, learning in the flow of work, which I spent a lot of time studying microlearning. And all of these phrases or concepts are iterations on the traditional paradigm of teaching.
[00:05:42] And I don't think that's really the right paradigm. I think the paradigm of learning is the learner is trying to consume or find or search or is curious. We're not teaching them, we are enabling them to learn and do their jobs better. And when we have a system that can dynamically build content and dynamically Translate it into a different language and dynamically produce it in a form that somebody can consume on their phone or at their desk or in some other form, we can enable them to learn and grow.
[00:06:17] And if you think about your life or your children or your grandchildren, the way they learn, yes, they go to course, a class and of course they need formal education.
[00:06:27] But for the first, at least three or four years of their life, they learn how to read, how to talk, how to walk, how to avoid being hurt, many, many things just through their own curiosity.
[00:06:40] And that's what happens at work.
[00:06:43] So I'm not saying that formal training is going away, but I would not be surprised if it's less than 20% of the actual learning needed in organizations.
[00:06:53] Now, as we studied this and talked to many companies and we've now implemented AI native learning in one of the largest pharmaceutical companies, one of the largest insurance companies, in a large airline, and some other big companies. So we know what's going on here.
[00:07:07] As you look at this, you realize that a lot of the infrastructure we've built in corporate training can now be automated.
[00:07:18] For example, one of the big trends over the last decade is role based learning or career pathways, where you go into a learning management system and you tell it your job and it says, here's the next set of courses, you need to get to the next level. And you try to traverse your career through this formalized learning process.
[00:07:42] Well, I hate to tell you this, I think that's ridiculous.
[00:07:45] That's not the way careers work. There are certain things you have to know and certain compliance rules and safety rules you have to know in certain environments. But most of us get promoted because of our experience, because of our judgment, because of the people we know, because of the breadth and depth of our T shaped development team. Meeting some depth, some breadth, our full stack knowledge. And we get that through exploration and experience, not through courseware. So. And by the way, building those career pathways is very expensive and it's very brittle.
[00:08:22] As soon as you build it, it's probably out of date. Because even though those levels or career paths are defined today, six months or a year from now, they might be different, especially with all the AI transformation going on.
[00:08:35] So that's number one, that all can be done automatically. Number two, skills taxonomies.
[00:08:41] This has not stopped, but it will continue. This idea that we're going to use the word skills to define everything that's going on in our company, I have a problem with it. I think it's very limiting in its concept, but it is Popular. And the idea here is we're going to take a list of words which we call skills and probably come up with 500 of them and give all of the employees a set of skills that they need to either learn or tag on their profiles. And then they can go into that skill and they can determine different levels of competency to learn that skill. Okay, let's take AI as a skill.
[00:09:19] What are you talking about here? Are you talking about developing AI? Are you talking about using AI? Are you talking about using AI in marketing? Are you talking about using AI to build social media marketing? Are you talking about using AI to build applications?
[00:09:34] There's hundreds of iterations and permutations of the word skill.
[00:09:40] So if you try to do this by hand, you'll end up with a simplified structure. And we always advise people to do this in a simple way. So you get the high level concepts, but you'll never get enough detail and it will never be up to date. Well, the AI native learning does this automatically. In Galileo, there's about, I don't know, 75 skills that we've defined. And every time we add new content, it tags it automatically by skill. So if you happen to want to learn that skill, you can click on that skill and look at all the content that's available and it will assess your level of skill from your activity and your behavior in the system.
[00:10:18] And if that skill changes because new content is discussing that skill in a new way, the system knows that you don't have to go back and manually tag things.
[00:10:27] And that gets me to number three, all the manual effort that goes on in training, in putting metadata around content for the lms, putting a scorm wrapper about around traditional documents and other content so we can track it, translating things into different languages, translating things into different domains of the company, all of those things are gone in AI native learning. Now, as we, as we did this research and Jordan, our lead analyst in this area, did a lot of the interviews here and I listened to many of them and talked to many of you about it, as did the rest of us.
[00:11:02] What we were observing was that this is not faster and more cost effective training.
[00:11:09] This is a different functional solution in business.
[00:11:13] And the reason we came up with the idea that dynamic enablement is because we already have groups of people in companies called enablement, especially in sales, in sales, which is, I think, one of the most dynamic roles in a company because you're dealing with the changing customer environment, the changing competitive environment, the changing economy, the changing product product, changing products. Changing prices, et cetera. We don't always call it sales training, we call it sales enablement. Because salespeople are on the job, doing things, talking to people, and they need to be enabled to do their job better. And I think that word applies to everybody. Sales, marketing, finance, operations, hr, all of us.
[00:11:56] We want to be enabled when the information and the skills and the technologies and the tools and the policies and the regulations that we need at a particular point in time. And sometimes we need an injection from the company that is formally pushed in our direction. But that is not the only approach. Most of it is pull.
[00:12:17] And so in the world of thinking about learning as enabling, we have to think of L and D or whatever you call this new function, as a much less centralized approach.
[00:12:29] So what we discovered is something I've been studying for three decades, is that the operating model of training, the operating model of learning has to be completely flipped to a distributed operating model.
[00:12:42] So what we're talking about in this research is most of us have corporate universities, corporate academies, leadership development academies, et cetera. And those are sometimes physical places where you come together to meet people, which is a great thing to do. And they're usually filled with massive course libraries of content.
[00:13:03] And the company buys this formally designed content from LinkedIn or Skillsoft or Pluralsight or Coursera, whoever it may be. And that centralized resource is available to employees all over the company to use. And that's all fine. I think the ROI of that is relatively low, by the way, but people feel obligated to do that to just meet the needs of various people in the company.
[00:13:29] But there is no way a central learning department has any idea what's going on in the finance department or the high net worth department in Singapore if they're located in London, or the high net worth organization in Italy or the Middle east, or to say nothing of other business functions around the world. All of those local groups, whether they be manufacturing plants, hospitals, retail stores, sales organizations, functional organizations have unique needs based on where they are and what's going on in their environment.
[00:14:09] And as much as you may try to understand that centrally, there is no way you're going to keep up.
[00:14:16] So what we do traditionally is we federate the learning function.
[00:14:21] We have sales universities, sales training, sales enablement groups, we have manufacturing training. Many times it's in each plant, we have hospital centric training in each facility, et cetera. Sometimes those people report up to corporates, sometimes those people report to the local group. And what we end up with is a hodgepodge of Content and materials, which costs usually three to four times more than the cost of the centralized learning function.
[00:14:51] We call that forensic accounting. In the old days we used to do this at Deloitte, where companies really don't know how much they're spending on training because it's happening all over the place.
[00:15:01] So of course we want to get our arms wrapped around that and we try to control it.
[00:15:05] Well, in the world of dynamic enablement, this can be done at scale because an AI native platform, which is consumed or filled with content from corporate or third parties, can be delegated to a local enablement group to add their local content. And so people in that location or that facility can learn things that are local to them, but also see what's available or needed or standardized at the corporate level. To give you an example, Galileo Learn, our version of this includes 7 or 840 formal AI created learning material or objects or courses on things that have to do with management and leadership and HR and recruiting and pay and diversity and career development and technology and all that stuff.
[00:15:58] We now have customers, more than 10,000 actually, who are using Galileo learn to learn about HR, but they're also learning about things in their own company. They're putting their own content into it and complementing it with our content.
[00:16:14] And that's what you can do in an enablement system, is you can complement a corporate resource with a local resource in an integrated way. It's very hard and nearly impossible to do that with traditional LMSs.
[00:16:27] Now, as I talked about briefly in the article here and you can read about in the research, LMSs are slowly going to die away. They're not going to disappear because we still need them for a lot of compliance reasons. But the LXP market is probably going to die off from this.
[00:16:43] Content development tools will be completely subsumed by AI native systems, and the way we think about instructional design will change. You can do instructional design by prompting now without having to write, you know, read a tomb of books on how to do various forms of instructional design. And as the instructional design and educational industry becomes more integrated and more involved in AI, there will be more and more instructional rubrics and paradigms loaded into the AI native system. Now, I don't think any of the AI labs are doing a lot of work on learning design. I think OpenAI has done some and they're mostly partnering with Courseras from what I can tell. But there will be new paradigms created.
[00:17:28] I think a good one is the notebook LM that came from Google. It's a spectacular way to learn.
[00:17:34] You put information into it and it creates a podcast. It's very easy to understand. You can give it a, I don't know, 200 page report. I've done this. And it'll create you a 10 minute podcast that summarizes it in a very compelling, interesting way. And that's early stage here. So things like that are going to be extremely valuable in your organization.
[00:17:54] The benefits of all this are very, very extensive. If you look at the research we're publishing, we've, we've looked at companies at the different levels of maturity here.
[00:18:04] And the ones at the top that are doing dynamic enablement are 16 times more likely to be innovation and market leaders in their industry.
[00:18:15] It's only 4 or 5% of the companies we've studied, but they're really pioneering companies. One of the examples we talk a lot about in there is Databricks. The head of learning at Databricks told me that she refuses to be part of hr. She thinks this is so important that she wants to work for the CEO.
[00:18:32] So the idea of enablement is bigger than learning. It's much, much bigger than learning. And so I think the CLO title, CLO Functions, CLO Operation really needs to be reframed as this technology reaches the market. The vendor market is beginning to adapt to this slowly.
[00:18:51] I think it's going to happen at an accelerating rate. We're certainly going to do our best to push it as fast as we can.
[00:18:57] There's a lot of new technologies moving in this direction. I mentioned a lot of the vendors in the report, or rather in the article.
[00:19:04] And if you want to be involved in this with us, let me just tell you how to learn more.
[00:19:09] 1. Get Galileo. Galileo has all this research. You can download the research report, you can ask it to benchmark your organization against the maturity model. You can look for case studies, you can learn about vendors, you can do what ifs, you can do financial scenarios we have now worked with. The clients we've worked with have saved 30 to 40% of their spending on L and D already by adding a much higher scalable solution and approach to L and D here. And so use Galileo. Second thing you can do is call us. We have a CLO forum that meets every month or two where we discuss this. We're going to be significantly demonstrating and highlighting this in case studies at our conference in June. Irresistible 2026. You can come there or we will just come out and visit with you and help you think through your strategy and put together a plan.
[00:20:06] If you're a vendor, if you're a content provider, if you're a content company, talk to us. We will help you think through your strategy. Relative to this, there's lots of opportunity for value creation in this new world.
[00:20:18] Selling traditional courseware may become much more commodity like or less useful. Your pricing might be struggling because of this, but it's a huge, huge new trend.
[00:20:29] We are extremely excited about it. I think it's a trillion dollar market. If you added up all of the implications on various forms of IT systems and knowledge management systems. It touches the world of digital twins which I've talked about before, which we can show you and I think it's a significant C level initiative. I would not delegate this entire topic to the head of training. I think it needs to be bigger than that because it impacts your skills strategy, it impacts your career development and career pathway strategy, your retention strategy and your overall focus on AI transformation and productivity because this technology facilitates the understanding, fluency and adoption of AI. We are really excited about this and please take a look at the research and call us if you'd like any help. That's it for now. See you guys again soon.
