LH128 Crossing the Divide with Lars Hyland

now we have these tools, AI makes it much easier to produce any content, but if it's not been well designed, it doesn't resonate at the other end. Just because you can produce more of that, so you can reduce the costs of production down to pretty much near zero, but really, as I called it, it's just faster GI-GO, you know, faster garbage in, garbage out.

If you're in L&D, Learning Technologies UK 2026 belongs in your diary. 29th and 30th of April this year, it's the UK's leading event for technology-supported learning, co-located with HR Technologies UK and bringing together 15,000 learning and HR professionals. There's a free exhibition with more than 200 exhibitors and 200 seminar sessions alongside a dedicated conference for L&D leaders. The focus is practical insight, peer learning, and examples from organizations including Google, Intel, and Merck. Bookings and registration are open now at learningtechnologies.co.uk.

This podcast is supported by Synthesia. Training videos shouldn't take weeks to make. Synthesia lets you turn text into professional multilingual training videos in minutes. It's fast, scalable, and so simple anyone can use it. You can write an email, you can make a video.

Hey there, My name is Alex, and I'll be guiding you through this training session.

70% of Fortune 100 companies use Synthesia to modernize onboarding, simplify compliance, and scale global training. Podcast listeners get an exclusive 30-day free trial. Go to synthesia.io/greatminds.

Don't miss it.

This offer won't last. Welcome to The Learning Hack, a podcast about the people and technologies that are creating the future of learning.

I'm John Helmer.

Now gets worse. Learning is cool.

Learning is cool. Learning is cool! I'm learning! Learning is fun! Knowledge is power! Knowledge! Education! Learning!

The e-learning industry has been selling something for 30-odd years. Content. Lots of it. Packaged, tracked, delivered, ticked off. And the pitch has barely changed. Here's the course, do the course, job done. What's changed is the scale. AI can now produce in minutes what used to take weeks, and a lot of the industry is treating that as good news. More content, faster, cheaper. My guest today calls that faster garbage in, garbage out. He's been making roughly the same argument since 2010 when he founded a company to tackle the gap between completing training and actually changing how you work. That company was 15 years ahead of its time. He's been at the sharp end of the industry ever since, nearly a decade running EMEA for a major LMS company, now building an AI-powered platform for small businesses who can't afford to get compliance wrong.

Yes, we go back, me and Lars.

We were both at Epic in the early 2000s when the e-learning model that he now argues against was taking shape around us.

Hack Facts.

Lars Hyland is Chief Product Officer at Enlightening. An AI-powered compliance and performance support platform for UK small businesses, and an independent advisor to edtech, learntech, and HR tech startups on product strategy and growth. He holds a BSc in Information Technology from the University of Salford and an MSc in Music Technology from the University of York, a combination that gave him an early research-grounded interest in how people learn through technology. His career has taken him from the early days of interactive multimedia through a decade as a senior leader at Totara Learning, where he built and led the company's global partner network, to founding Retenda in 2010, a learning reinforcement platform built on spaced repetition a decade before it became fashionable.

In our conversation, Lars takes a pretty bleak view of how AI is being deployed in L&D right now, but he's not pessimistic. He thinks the tools are forcing the industry towards a question it's been avoiding for years. And that question is a good one. We also got into the LMS wars. Lars has spent nearly a decade championing open source learning infrastructure while simultaneously acknowledging that most LMSs were doing exactly what the critics said, measuring activity rather than changing behavior. He has a view on how you square that. And because we go back a way, I pushed him on the bit that doesn't quite add up. If you believed all this since the early 2000s and the industry still defaults to tick box training, What exactly is going to be different this time?

So Lars, it's brilliant to have you on, and I should flag for the listeners that we go back a long way. You're already a fixture at Epic when I rocked up there as head marketing Muppet in the early 2000s. Though this is a slightly unusual one for me. Welcome to The Learning Hack.

Thank you for having me, John. It's a delight to have a chance to reminisce with you and to look into the crystal ball.

Well, not so much looking at the future, there'll be a bit of that, but there'll be quite a lot of looking back at the past. So I want to cover your really interesting career Lars Hyland, this is your life.

John, thank you. I hope it's not the end yet, you know, so I think there's a bit more to go.

Yeah, so when people win a lifetime award, not that this is the equivalent of a lifetime award, but when people win a lifetime award for achievement, you meet them afterwards and the first thing they always say is, oh God, does that mean it's over?

There you go.

Tough luck, David James. So the epic days, you were doing very well there, am I right? Big retail banking clients hitting your numbers. Looking back from where you are now, what do you think we were actually selling them? And did it deliver?

Well, that's a good question because we're going back quite a while in time, aren't we?

Yeah.

Early 2000s. Early 2000s. Well, you know, it's probably best to go back even earlier than that because, you know, I started my career at the beginning of the '90s, really. So I got into what was called multimedia or hypermedia, even if these terms still even exist and resonate with people. So it's pre-browser, so pre-internet and using interactive video discs. So that was the sort of basis on which I cut my teeth as a training designer. And, you know, I arrived at Epic at, in '92, '93, having done an earlier stint at an organization that's worth shouting out, Future Media, who were a pioneer at the time.

I later worked for them.

You did indeed. So they were a pioneer at the time based on some good research that the founder there had done on accelerated learning, which was also what we were all talking about back then if you were delving into trying to work out what would work best to support people's ability to acquire skills and capabilities. And a lot of that research is still extremely valid and neglected. Well, it's more known now. I mean, we're rolling forward a few decades now. But I just wanted to sort of COVID you know, how did I even get into this sort of area? And the reason was, is I was actually studying a music technology master's. So that's another commonality you and I have, John, right? But I was very much into the synthesizers and the samplers and strange noises. But actually what I did on that course was quite a bit of hardcore coding. Messing around with MIDI data, but also sampling and software synthesis. And I realized that I was the only one creating programs and putting a help function on it. So I actually tried to help people work out what to do with fairly complex tools. And then I sort of investigated that area further and I realized that musicians are fantastically good early adopters of technology when you think about it. And I then got further into the idea of, well, I'm building this software, which is extremely flexible, but also undercooked. If you think about the computing power of a basic PC back then, to what you can do now on your, literally on your phone locally. You know, we were doing some quite interesting stuff. But it was quite complex. It was literally asking musicians to sort of play with C functions, which in a coding language, which is not ideal. But if you put a GUI on it and could explain what it was doing, I realized that if you embedded that around the actual tool that created the sounds, then people could get much more out of it. So this is what got me into hypermedia, right? And hypertext and creating an interactive environment. So that's, that's so, so, so So my epiphany there was, wouldn't it be good if everything, every tool that we use, every environment we have had support there with you, which as it turns out is still a novel idea all these decades later, although it's now much more prevalent and thought about that we can embed in a technology sense performance support. I'm sure we'll come onto that in a bit. Just to loop back to your question about, you know, joining organizations.

So before you go there, Lars, I think we ought to flag up as well that it's keyboard players who are early adopters of technology. Guitarists are still using 100-year-old technology, you know, guitars designed in the, you know, '40s and '50s, electric guitars, and their amplifiers are full of valves, which is very early and still, you know, But carry on. Let's get back to that question. How did you feel about what we were selling them in those days at Epic?

Yeah.

Yeah. Well, I was very, I guess, naturally consultative because I had an actual interest in the actual underbelly of how do people learn more effectively? How do you train people at scale? And I think that's the critical word here, at scale, because an organization like Epic, we would go out and we would be supporting large corporates, and those corporates had big change programs, and those change programs involved typically thousands of people that needed to start doing things differently, either through being compelled to because it was compliance and regulatory driven, or it was a big system change in a way their actual everyday ways of working were going to fundamentally shift and change. So, so the attraction of e-learning or interactive training back then was that particularly networked. So we were talking about that when it became networked, that you could do these things cheaply at scale. I mean, that said, we did start with CD-ROMs, if you remember that, John, you know, we were sort of building those and CD-ROMs and then being able to publish them out and then get them out to people's desks. And then they could work through simulations essentially of the working environment that they were now needing to engage with. And then that simulation would give them the necessary sort of support. So in that context, I think those solutions were absolutely necessary and did a lot of good and help. I mean, I think we then moved into the era of, with the e-learning module and basically the internet and interconnectivity coming in that allowed you to ditch the physical stuff. But actually we went through a period of kind constrained bandwidth of video quality, audio quality to a large extent, and replace that with the fact that you could connect everybody to something very easily and cheaply. So it's about reach and scale. But it did compromise, I think, in the early days at least, it compromised the the actual experience you got because it got reduced down to a sequence of interactive elements with some nice graphics in, if you're lucky, and a quiz. So, and then we sort of now, we still live with that now, you know, that we sort of, that de facto sort of approaches which the team at Epic and the early pioneer companies really were setting out the actual underlying operational production practices to sort of build that out.

We were learning as we went along.

So I think in the early days, the early, early days, we did actually focus on effectiveness because we had the technology was more isolated and primitive in the sense, let's say primitive in terms of the computing side of it. But video, interactive video, high bandwidth video, we were creating very sophisticated simulations of interviews or environments back then.

Oh yes.

I don't know if you remember that.

Yeah. Yeah. SAP piece, which never got used, I think.

Yes, maybe. But there were plenty of other solutions which I did in other— prior to arriving to Epic at that time, which were used in organizations who had the money and, if you like, the resources and space to sort of deliver that effective— that type of technology at the time. But I think over time, obviously, yeah, it got replaced. Effectiveness moved to efficiency at scale, I think. And we've been in that era for quite some time. So how did I feel about that? Well, I think it's important to say that whenever we were out there talking to customers, our team then, were thinking about the more holistic sort of learning solution that you'd need. And that would often be pitched, right? That was what you'd pitch. But what got bought was perhaps not all of that solution. And that just came down to often the right solution that would deliver the maximum impact takes more time and more, bit more budget than organizations were willing to stomach at the time. And, you know, I think that's still often the case.

So that was the era, of course, I suppose the early 2000s where the model kind of solidified around, you know, LMS plus content, to put it crudely, that I was talking about there. It is worth saying that it's not like you were uneasy about that, but the rest of the organization was 100% with it. Because as you say, I mean, Epic was really quite forward-looking in the things it recommended. And of course, Donald Clark, CEO at the time, was cranking out white papers, you know, one a week, arguing against that model for the most part. Yeah. Which it was continued to do. So was there a moment where you realized that was happening, found it was difficult to get clients to look beyond that, and felt uneasy about things? Is there any particular moment that shifted you?

No, I mean, I think it was more, I was always aware, always very much aware, just because I was naturally researching the whole field of what does good learning design look like and what does that experience need to be? And it, you know, as we all know, well, I say all, you know, it should be more widely known, but, you know, a good learning experience isn't just content. It's not just consuming one item or, and expecting people to be miraculously changed at the end of that, you know, the old Cheek Dip, which was a common phrase that we used even back then. It still happens a lot today. It needs to be a much more considered blend of learning activities and practice and reflection activities. So there's plenty of good practice that's been written about that, but it hasn't necessarily become the de facto approach to when applying it. That said, I think it's getting much easier to sort of shift to that model, which I can talk about. But I was aware of it back then. I would be presenting that to customers. Some would would buy into it. And others, as I say, it would default back to the sort of the core, right, we need a bunch of people to get through these courses by this date. And that was the sign of success for them.

So in 2010, you founded Retenda, which is in a way a reaction to that, a reinforcement learning platform. Built on the idea that training transfer can be as low as 20%. So sheepdick training is probably losing about 80% of what you're trying to teach people. That's a problem to build a company around. Where did the conviction come from?

Again, delving back in time, you mentioned how guitarists still use old technology. Well, you know, you delve back into the archives of research and you can come across figures like Ebbinghaus, you know, back in the late '80s, looking at memory and reinforcement.

Well, I have another podcast where we delve heavily into the—

I'm sure you do. Yes, exactly right. So there's that, but then building further onto more practical aspects, there's been research done around interval-based reinforcement, spaced learning, as some people would call it, and that's become much more recognized now and it's either a feature or it's a focus. There's good solutions out there then, now. And I was ahead of the curve, I think, in identifying that. We built Retenda and it was all about shifting the focus to support people's ability to be nudged, essentially. So again, use the word nudge. We didn't use it back then, but if behavioural sort of techniques to sort of nudge people to start using new knowledge and skills at an appropriate moment. It's only then that you start to shift that into your sort of personal skill box, shall we say, and it becomes a skill and capability that you use. So yeah, we built that. We had some good traction with some international training companies using it. So Capita and Bupa at the time were customers, which was great. But one of the lessons you learn, that you can come up with great innovation, right? Again, in business, a lot of things is about timing as much as anything else. And so Retenda was a lot of time spent educating the market to sort of recognize the challenge and the problem. People would recognize the challenge and the problem. But like I was saying earlier, the sort of gravitational pull of saying, well, okay, we need a, you know, we need content in front of people. We need people to complete that content and tick a box. And then we run out of time and effort or the agenda moves somewhere else. So the desire to sort of really follow through and genuinely build capability and skills, you know, you really need to find people who sort of understood that. So, so yeah, that's a great concept. And if you look at it, look out in the market now, there is a category, you know, there is, there are people out there who deliver specifically that type of solution. And in fact, actually, it's becoming more and more subsumed, I think, into just into as a feature set, you know, LMSs, but also, um, even actually even going beyond that now as we, as we march forward into AI support, um, then it just, it just becomes a natural dynamic if, if you direct it in the right way, um, as part of that experience.

So maybe Retenda a bit too far ahead of the time. Might be easier, right, if you started it today. I mean, you know, it's not only having a great idea, as you say, it's the timing.

Yeah.

So after that, you spent nearly a decade leading EMEA for Totara, an LMS company, at a time when the LMS was starting to get a fairly sustained kicking from a lot of progressive voices in L&D. Still is. I mean, you know, every day you can open up LinkedIn and see someone else heralding the death of the LMS. And I know you'd probably agree with all that. So how did you square that, that you were there in charge of an LMS company? Well, from EMEA anyway.

Well, that was actually part of the global leadership team.

Yes.

And it was a global, and is a global organization. Well, I did spend a short period of time at at the bleeding edge, I suppose. So Retenda being an example of being at the bleeding edge of some innovation and trying to bring that to market. Did the same thing with an adaptive learning technology, which was an early use of AI as well. And realized that while we were doing some good things with small numbers of people in terms of impact, actual impact, I realized it would be great to work in an organization where there was much more scope to go much wider. And Totra at the time was doing well, bringing open source, an open source solution to the market. And it was timely to sort of come in and help build out what was the sort of front of house of Totra., and, and manage a global network and grow a global network of partners, around the world. It was quite a different way, go-to-market model really, to, to, to the way even now, you know, and, over that decade period, you know, we got to the place where we had about 20 million-odd, end users, half the NHS, I think they still get their learning through it.

Which is a very big organization. That's one of the biggest.

1.4 million people. But also, you know, it was a proud moment. It was a sobering moment, but a proud moment during the pandemic that we were able to sort of support the sort of Nightingale hospitals going up in the pandemic. Wow. And also in Italy as well. That was quite sobering. If you remember, they were badly hit early on and they were, you know, their healthcare institute was using using Totara, and we were able to sort of just enable them to go from, I think it was 40,000 healthcare workers, sizable number, to 400,000 in a month, um, to sort of get them up to speed on ventilator training and other things that you just needed in an emergency. So, so, so the reason I say all of that is because, you know, a platform like, like an early management system, and if it's reaching, you know, at a global level, and you can start to bring, as we did over time, sort of shaping the sort of feature set to sort of lean towards the opportunity to build more holistic, more blended, more effective learning solutions as part of that core platform. That's going to, you know, that aligned, if you like, with my thinking to sort of, you know, how do we embed effectiveness? And not just scale.

If you're looking to connect with the top minds in learning and talent, there's no better place than an iVentiv event. Since 2009, iVentiv has hosted over 300 events for global heads of L&D and talent. I've attended multiple iVentiv events myself and found them to be invaluable, less about sitting and listening and more about real impactful conversations. It's an unmatched opportunity to collaborate with leading chief learning officers on the challenges they're tackling today. Join the conversation. Visit iventiv.com/learninghack. That's iventiv.com/learninghack.

Open source was a big thing back then. It, I mean, it's still around. Some of the AI models are open source, for instance. And I'm interested in, curious about, I suppose, what it's like, how the open source model, which it was supposed to be about changing the power relationship between vendors, developers, and buyers in one sense. When you have an open source technology and it then becomes a kind of commercial entity like Tara with that kind of different kind of dynamic. Does the open source element of it still continue to make it something different or does it change completely? And now it's a commercial beast. Now we, those things are over. We're interested in what happens there and you know, what your reflections are on that.

Well, I mean, it's, I mean, I can speak more generically really about the use of open source. I mean, open source has been around for, a long period of time. In fact, Todra came into being because it built on top of Moodle, right? So Moodle is probably the most successful, sustained, community-led sort of build-out of open source, certainly in the education area, sector, tech. And that's also gone through its own sort of journey of being a fully community-built or supported and led and built solution, which has huge traction because it really spoke to a need.

Yeah.

And then Totra came about because that need was also there in the workplace, not just in the specific education context. So it was a good platform on which to build. And in fact, that's the whole point really of open source is that you can stand on the shoulders of giants and sort of help keep building things out at a pace that you find difficult if you're trying to do it just unilaterally as one organization with a limited pool of resources. So conceptually, if you think about it, pretty much every technology we're all using now has got elements of open source. The whole internet is built on open source software. So it's always a mixed model. You see, you've got full community-built open source. A lot of those projects sort of start to fray at the edges as the community sort of gets interested in other things, or it gets adopted and then turned into a sort of mixed model, which you can see, and, or it gets, you know, you switch over, you have a full proprietary type of solution.

Yeah. But is your, the kind of impetus of innovation in that product, when it becomes a product, does that become very different? Does it slow down? Is it harder to get the innovation going?

I don't think so. I don't think so.

When you have more limited open sources.

Well, I mean, no, I think it doesn't relate to open source or closed source. I think it's, if it's managed well, then you can have significant acceleration of production. But in a way, there's a bit of a moot point now because we're now in an era where we can tap into sort of code production using AI and AI agent support, which opens a massive opportunity for, you know, well, either experienced or focused individuals or small teams to build at scale and at massive pace. And we're, you know, so I think that's another tranche here of— of what we're going to see really manifest itself over the next 2, 3 years is how new technologies, new businesses can be built with small teams and sustained as a small team and be building out the necessary feature sets and improvements to a platform using technology, building technology.

So that brings us up to date in a sense, because you're now at Enlightening, which is kind of about compliance and focusing on SMEs and so on. But tell us about Enlightening and what you're doing there.

Sure, sure. So this is a bit different and it's one of the things I've been delighted to sort of help bring to market. And it's a different sort of area. So while My experience largely to that point has been about enterprise, large organizations, how do you support learning and development at scale and deliver change at scale? I was always curious about smaller organizations and how they get supported. And often when you're very small, then you don't have the infrastructure, you don't have the resources either to sort of do things in ways that you'd like necessarily. And if you boil things right down to their sort of core, then if you want to grow a business as a founder or as a leader in an organization, then you need infrastructure, you need structure, you need to be able to document that. And then you need to have a means by which your team or teams can operate to those standards. So that's really what Enlightening is about, is to sort of help people cross a Rubicon or a chasm, should we say of, well, I've got to this point through informal communication because I can, because I've got a small team, but it's exhausting as a founder. You're wearing all the hats. How do you write that down effectively in a constructed way? How do you meet compliance, external regulatory demands on you as you try and build a business that serves bigger businesses. And often people struggle with that. You know, they, they, they sort of hit a ceiling and can't get past that. So what we've done is built a platform that really streamlines the ability to create an organized set of standards. So that includes your compliance, the regulatory stuff that you need, whether it's just, whether it's company policy, HR policy, or sector-specific. Materials, and then using the power of AI to sort of personalize and customize that to your particular needs. And then automagically being able to produce very targeted, contextualized learning performance support that can be, that can be really tailored down to even the individual in your team. But the beauty of this system is that you remove a lot of the administrative of overhead that stopped people going there or even going and trying to do this. But now you have a mechanism by which you're able to provide consistent advice, 24/7 support, because your team can talk to these documents and can be guided with very specific, um, you know, what would you do in these situations? They're literally the situations that you find yourself in, in your role in that organization. So it's much more much more actionable. So that's the whole principle. So it's moving away from packaged content to a flow of context. That's one way of putting it. So yeah, so it's much more about that ongoing support and getting to a place where leaders free up their time. They're no longer answering the same questions all the time, trying to struggling just to keep their team aligned. They're the human in the loop trying to do that. And to sort of take 80% of that away so that people can get the answers when they need it, build their confidence and act, but then to surface those opportunities to improve. So you start to genuinely spot the edge cases. You have the time to deal with the edge cases. You have the means by which to quickly capture and document those edge cases. So then you get this improvement loop going, which is often missing. So that's what we've got. So Lightning is aimed at small businesses. It's a cost-effective solution to really help those organizations get structured and put themselves into a position of growth. And when it comes down to it, they're a business. They need to be able to get to a place where they're investable if they want to keep growing. Right.

So you're focused on compliance there, which is kind of, you know, has always been the sweet spot for e-learning actually, because it's something that everybody has to do. They have to do it at scale. Nobody gets all that excited about it, but it kind of has to be done. Many, many places it's done in a tick box, others it's not. Certain industries put a lot of effort into, you know, if you're in oil and gas doing health and safety, for instance, it matters an awful lot because, you know, people die. So I can see the kind of commercial logic there and it makes absolute sense. And you've given a really good explanation of what you're actually doing now, which I'm ashamed to say I didn't really quite grasp before this interview. I think it's actually quite— it's an exciting idea, I think. But I wonder if it has a wider applicability beyond compliance because— and the reason I say this is yesterday I was at a podcasting and learning exhibition. And I was telling people that if they want to move from just asking questions to an LLM to being a bit more agentic and getting AI to do more within their workflow, what they do is need to document their processes. I am aware that small companies, this does become a barrier because, you know, when it's that kind of classic founder situation of you know, maybe 12 people in a room, 25 people in a room, whatever, and you can all talk to each other very easily. A lot is kind of tacit, understood. As the organization scales, you get this massive problem of having the time to put all that stuff down and make it explicit. And as you say, put structure in there. Do you think there's a lot of scope beyond compliance? Is that part of your kind of It is, it is. Day 2 planning, your day 2 planning.

It's day 1 planning. It's part of it because I mean, you're right. We currently lead with, it's about compliance because that's the element that people have to do and they, you know, there's a pain point there for every organization to sort of be operating safely. But really, you know, I use the language of, you know, of having your team operating to the standards that you set, which is a slightly broader statement if you think about it. So by those standards include compliance with regulatory external sort of stuff placed on you, whatever your chosen sector you're operating in. But then there's how you want to engage with the market. You know, what is it? You know, how are you presenting yourself? And then what is your product? What is your service? How does that interact with itself? And that's the sort of thing that you do need sort of document. So your operating procedures is one way of putting it. So all of those controlled documents, that's an even broader but dull statement, a phrase for it, is how do you manage that effectively? Actually, we're sort of talking about knowledge management as a phrase, right?

Kind of. Yeah.

Well, very much so. I think, you know, it is knowledge management. It's the sort of blueprint by which you keep you have to keep up to date to define what you're doing as an organization. If you don't keep it up to date, then people don't know what they're meant to do. And so they're not aligned. And I think if you can envision a situation where a lot of the perceived busy work and real busy work of keeping on top of these documents is actually, you're highly supported, so you reduce that down by 80, 90% of the time it takes. You can move into the better place of always looking for opportunities to improve. And that accelerates your innovation. It means that your people always feel like they're engaged. They know what they're meant to be doing because they can trust the sort of support they're getting, both the automated support and also they can spot opportunities to improve. And then you've opened up time to do that. And if you do that, then you start to become more competitive. You're going to grow. And if you're in that situation, and any founders who are listening will be conscious of this, if you're building a business at some point, how do you realize the value of that business? And if that's involving external investment, then You need to be investable. To be investable, you need to have this architecture in place. You know, that's attractive. It means you're more valuable. It means that you can be picked up and scaled, right? With some external support in a much more managed way. And that's obviously, you know, that's a big motivator for smaller businesses.

That's really interesting. I mean, it's interesting to me because that thing of crossing the chasm, I've been very aware of this, that statistics will show that there are kind of breakpoints in the developments of companies. Some of them, many of them, and you'll know many of these because there are loads of them in Brighton. Many companies never get beyond 25 people in a room, maybe 4 million turnover a year if they're lucky.— we're both in companies like that, and make it across to more like kind of epic learning pull scale and beyond. And I've always approached that from a kind of marketing point of view because in my consultancy work, because that's kind of what I do. But one thing I've become aware of in that is that with learning particularly, that part of looking at that breakpoints and how organizations evolve is when they get typically they'll get an HR manager first, and that person will have some responsibility for training. As they get a bit larger, they might get a kind of learning person as well. It's quite a while before they get to the kind of edifice we have at the moment. And we know that the number of HR people in companies, US and UK, possibly Europe as well, has dramatically increased over the last few years and is getting bigger and bigger. Does this go beyond learning to HR as well, do you think?

I think so. I mean, I think that's an interesting sort of trend, isn't it, which might change. I mean, I was going to make one comment, which is if you look at the direction of travel, scaling a business, There's a potential decoupling, particularly if you're a technology-oriented business. There's a decoupling, scaling revenue and scaling impact in getting your product and service sort of used out there and the number of people you need to sort of support that with the right approach to adopting sort of modern technology, obviously including AI and agentic sort of mechanisms, is that you need fewer people. So to your point about, say, here in Brighton, I have been in some sessions about future skills for the region, Sussex-wide, and also about positioning companies for growth. You're on the board of WIRE.

You hired Sussex, weren't you?

Yes, I was.

Which still owes me 60 quid, I might add.

Oh, does it? Oh gosh. Well, it wasn't specifically my fault, but I'll let you off. I'll buy you lunch.

Oh, a good lunch.

It would be a good lunch, wouldn't it?

Yeah, yeah.

Yeah, a place like Brighton, which is actually extremely creative, you know, it's often straddling the very interesting sort of areas of technology and like media and all sorts of different areas has struggled to scale in terms of organizing themselves and getting ready to sort of grow bigger and break out. But it's interesting because I think, you know, there's never been better to get to a point where you have 20, 25, maybe 30-odd people. What you can produce as that team and what you can do at scale now is amazing. You know, you don't need the complexity. You can dodge the complexity of having thousands of people potentially. I mean, obviously, if you get global, then that sort of changes a bit. But I think the support you need to sort of grow is, and to do that consistently. So the other point I wanted to say, you know, you mentioned about about having these roles as you grow. Well, if you get the core mechanism in place, like a tool like Enlightening can delay the need to have a human involved for longer, which actually saves you money and cost and also gives you a stronger foundation on which when you do bring someone in for them to do much more with that at scale. So it sort of helps you balance what you actually need in terms of resources., as you grow from being a, say, startup-level smaller team to that next level where, as you absolutely point out, you know, communication gets more difficult, alignment gets harder if you don't have these sorts of tools in place.

Yeah. Okay. So the big role for AI in this company, in Lightning, you've been critical of how the industry is using AI mostly to produce more content faster, which you think is the wrong application. You're not alone in that opinion. Other people argue against it and say, well, you know, we're seeing all kinds of other types of activity going on with AI and in L&D that aren't necessarily just that, even if that's the main thing that people are doing. But what does good use of AI in learning actually look like for you? And can you point to anyone doing it well?

Well, I can point to Enlightening. But, um, uh, I know, I mean, it's worth pointing out the context. I actually spoke about this in a conference conference in Warsaw in Poland just late last year. My talk was called The Emperor Has No Clothes, as in, and the context of this is it was firmly in the area of, well, and we touched on this already, is that the emphasis of creating content at scale and getting people to wash through that and complete it and that's job done in many cases, is that now we have these tools, AI makes it much easier to produce any content and push it out to people at scale. But if it's not been well designed and there's a lot of context that gets ignored when you build content because you're doing it generically, either because it's literally generic content that is being bought into your organization, or it's you build one piece that goes out to many, many hundreds of people or thousands if you're very large, it doesn't resonate at the other end. Just because you can produce more of that, so you can reduce the cost of production down to pretty much near zero. But really, as I called it, it's just faster GI-GO, faster garbage in, garbage out. So you've accelerated that. And I've put forward the question is that rather than, I mean, that's a negative statement to say, but actually there's a positive outcome. Positive outcome is if you follow that track, then eventually you're forced to confront the question of, well, we can produce content instantly, pretty much. But okay, it's still not having, it doesn't have any impact. So let's ask a better question. You know, what should we be doing? And it's clearly not just that content. You know, you need to be thinking about all the other things that need to wrap around any change requirement. And then you realize that it's not about, you know, a content sort of led experience. It's a whole suite of things that gives people the ability to put new knowledge and new skills into practice. And that's a different type of experience that's required. So I find that this, we're in a period of transition, I feel, where that realization comes about because these tools can allow you to design what was the harder, messier bits that people people, as I said earlier, didn't want to do because it felt longer and more difficult. But well, look, now you're free to just focus on that. And then when you've done the hard bit and defined what the small elements of highly contextualized content is that you need, brilliant. You've got the tools available for you to produce those things at a desirable quality level, very quickly at very low cost. But now that's a better way to think of it rather than doing what we've been doing already, but cheaper and faster because that's just, as I say, faster garbage in, garbage out.

So have you seen the green roots of people doing more useful things in any particular examples or are you just too busy building out your business to look at part of the big picture?

It's not, yeah, I mean, I am doing other elements as well. It's yes, yes, there's some good work going on in terms of the application of AI to in the more wider talent space. So there's organizations like Talent Mapper, who have tools that help get over other pain points around defining what the skills are. What skills do you have in your organization? This is for larger organizations, obviously. Well, it still applies to smaller ones, but what is the skills mix that you have in your team and teams and people? What's missing? What do you need to sort of help you develop going forward? And then how do you close that gap? So it's a more intelligent way to sort of identify Um, you know, the specific skills gaps that will make a positive difference to the competitiveness and performance of the organization as a whole.

But you're not naming names?

Well, TalentMapper is one of them.

Apart from TalentMapper, yeah.

Yeah, uh, there's another, um, there's other tools that actually focus in on L&D themselves, so the community that we're talking to now would be interested in tools like epiphany.ai, who are trying to sort of use AI to give an accelerated application of better learning design practice. That's not actually what they say on the website. That's a mouthful. But essentially, you know, all the good practice and all of that, and I would say is often not applied because it's not been easy to sort of tap into it. You can actually have highly trained AI sort of coaching and guidance to sort of help you get to that more considered design solution more rapidly and get it to a place that it can be built and made available to your people. It doesn't solve the latter part, which is the implementation and delivery of it. But that's again, an area where, you know, I mean, there's people like out there who are very much more than I would say predicting the of the death of the course. So the course goes away. And the way I've just described it, you know, that sort of simplistic, here's a course, do it, complete it, job done. That does need to go away because it doesn't really add much value other than in the context of a legal requirement in a compliance context, which is what keeps everything all the lights on at the minute. But if you move to that more performance support model, then you're going to be much more aligned to shifting actual skills capability.

These are mostly kind of vendor proposition examples you brought up rather than kind of practitioner sides. And I think that's understandable because of what you're doing and where you're positioned. Slight change of subject, but not a huge huge change, really. Something that seems to be very much under pressure is the SaaS model, which is software as a service model, which most of the bigger learn tech vendors now subsist on subscriptions, web-based applications and so on. Is this still the right model when organizations can increasingly run bespoke AI on their own infrastructure? I mean, since you spoke about that, in 2025 at LearningTech. We've had OpenClaw and the SaaS apocalypse, where, you know, which was something that happens in the stock market where SaaS stocks all went down. And so it's kind of a shudder through the learn tech world, I think, if nothing else. What does the world look like now for vendors who built their whole business around subscriptions? Should we be panicked? If we're in one of those businesses.

Yeah.

I mean, in many respects, it's not just— I know you've honed in on a SaaS model, but if you think about the implications of the broader implications of applying what is a highly scalable, intelligent resource, which is not a human resource now, it's an AI-oriented resource, and that is now available to you as a, and any organization in any field, then it gives you, I think you're right in that it gives you the opportunity to be able to build bespoke solutions as you go.. And if you're operating in an environment that's changing all the time, and you have the toolsets to meet that pace of change, which you can, then yeah, I think you have a situation where the whole— there are certain classes of packaged solutions out there that are SaaS-based or software-based. The value disappears. You won't pay as much for them. So I think it's more that. I think there are still going to be areas of the market that will still value and will prefer to do that and not do it themselves in that sense. But until the AI infrastructure sort of matures further, but I would say there's big chunks of that, but it's not just SaaS tech vendors 'cause it's, it may, many, maybe they're the canary in the mine, sort of thing.

Canary in the coal mine. Yeah.

But I think that all said, I think there's a huge wave of innovation that's flowing through that is all built on and around AI. And that we've yet to see the beginning. This is just beginning and we've yet to see the shape of how that operates. But you know, that level of, if you like, technology orchestration is something that a whole, you know, lots of businesses will not want to do that themselves. They'll want to tap into expertise that can do that. And it's just a, if you wanted to talk specifically about the commercial bit, which people are, you know, is it subscriptions? Is it now a sort of value-led you know, you pay for the value you're doing.

Well, things like the all-you-can-eat model with content are coming under pressure now. And people are looking at that thinking, you know, why do we have to buy all this content? And I'm now using AI to get rid of a third of all the courses I have because it's helping me too.

So to the earlier point, you're probably wise to get rid of all of them. You could make that, I mean, if you took it to the extreme of saying, well, well, look, I want a highly contextualized, personalized, yes, some content, but it's a focused thing for an individual. And naturally, the only way to do that is using a more conversational, adaptive sort of experience, right? So having packaged any type of packaged content doesn't really meet that particular need. You need to sort of constantly iterate it or have a means by which you can constantly iterate and contextualize and personalize that. And And I think that's the direction of travel is towards that type of experience because you don't have, you know, because you can tap straight back into the actual source of truth. What is that for your organization? Which then comes back to what we were saying earlier about knowledge management. If you put your energy and focus in defining that well, then then you don't need to sort of have this, which if you think about it, it's a bit of an artificial layer of training things abstractly, just in case, in many cases, which has then become something you have to maintain and pay for on a regular basis. Maybe you don't need it anymore. Maybe it needs to be something else, in which case that whole layer gets disintermediated. It goes.

It's so interesting because I think that in those huge catalogs that a lot of people maintain on their LMSs, there must be a lot of duplication. It's like, you know, we've got, we bought thousands of courses off, you know, Avenda. We've got quite a bit of a suite on leadership here. Oh, there's this new specialized leadership program that's just come out. We better have that as well. We're not going to get rid of the old one just yet. Oh, and yes, here's another new one. You must end up with a lot of stuff that kind of overlaps. So what you're kind of suggesting there is that maybe you would just have one way of teaching leadership, which you could continuously update.

I'm not saying there's one way. I think it's much more, yeah, okay. There's one way of saying it, which is in my organization, like I was saying before, what are the standards I'm setting? So one of the standards are going to be my approach to leadership. So, and there's different schools of thought arguably around.

Yeah, yeah.

Very different schools of thought. And so you can choose whatever Whatever one's resonating for you and you want to say, well, I'm going to adopt that. That's the framework for you. And then you need to hone it down to the individual person who's going to be developing their leadership skills and that person has some prior experience. They have a context in which they're operating in the organization, their role and where they're heading. And in the past, it's not been possible to think about, well, I'm just going to design a leadership course for John Helmer in this role, in this company, because that doesn't scale. It's prohibitive. Now you can. I think the opportunity is open for that. Obviously you're being guided now by, and you need to trust it, which is another question, but if you did, an AI sort of guide that works alongside some human support inevitably as well, giving you access to tasks, like real tasks that you do in the company to demonstrate these new leadership skills. See, that's the experience that's going to work and be more sustained and actually make a difference. And actually, you immediately have positive impact on the individual, the team you're operating in, and the organization as a whole, rather than, all right, you're going to go pack you off for 2 weeks on a leadership course. And yeah, I mean, I think most of the time those things don't really add a great deal of immediate value.

So many, it raises so many questions.

Yeah.

And that's always the case with really interesting guests is that you ask them a question and it just raises about a million more. Conscious of time. I'm also conscious that I've kind of given you a bit of a grilling. It does tend to be friends who get worse grillings than, you know, complete strangers. So I apologize for that. But I think it's yielded a really interesting discussion. For me. But finally, trying on a lighter note, you recently published a blog post about is your organization jazz or classical? Interesting for me because my daughter is about to graduate from studying jazz for 3 years at London Conservatoire. And so we have constant jazz in the house and I've learned a lot more about it. How much do you know about jazz? And what's on your desert island discs? Having started off with a kind of this is your life theme, and to see how kind of stuck we are in really old world legacy media formats, we're now going to end with desert island discs. What is on your jazz record shelf that you could recommend? And it might be a Spotify playlist rather than the shelf nowadays. Can you give us 2 or 3 things people should listen to that exemplify for you what jazz sounds like when it's doing its best work so that they can apply this to their working methods.

Sure. I will do. I love jazz, but I equally, as you know, John, I've got quite an eclectic taste music-wise. But just to allude back to that article and all that sort of line of thinking that I started quite a while ago, actually, I've used it a lot still. And in fact, it's now resonating even more was to put out the, you know, as if you're an organization So you're a group of people, basically. You know, typically as you grow, you become much more like an orchestra, you know. So yeah, standard operating procedures are still very important. And as I've just been saying, they need to be right and defined and iterate and evolve. And they become like your score. That's your, the conductor with a score and you have different people in different roles playing different instruments. And they all need to abide by that. That's still very very much part of how an organization needs to operate. However, the environment we're in is moving at pace, and not just pace, exponentially. That means, as you know, that just means everything's getting faster. People need to feel, not just feel, but be confident that they can literally improvise, right? They can need to improvise in their role. And that is actually the nature of what startup is like, because you you have the opposite, which you don't have anything guiding you necessarily. So, but actually jazz, you know, you do have some guiding principles. You know, each person does know their sort of role, but you, it's an absolutely built into this, into it that you will improvise. So that was my, that was the article. So I'm basically saying be more jazz. So it's basically have, you know, you have some structure, but free people to innovate around that and then have a feedback loop that will turn that into mechanisms that other people can then operate with.

Right.

What can I recommend in terms of— Yes. Well, I started my musical career as a clarinetist, so it was orchestral. I then got into Dixieland jazz and I was part of a little Dixieland Jazz troupe. So Louis Armstrong, all that lot, and jazz, um, started there. But then Webster Coltrane, Miles Davis, all really strong, uh, you know, like that.

Sydney Bechet on the way.

Sydney Bechet, absolutely, absolutely. And then more— well, not recently, this is still early, well, late '80s, early '90s— great British collective called Loose Tubes.

Loose Shoes. Oh yes, yes, know them well, kicked with them.

I used to—

yeah, when I was in a band, we, we, uh, shared a rehearsal space with them.

Well, there you go, excellent. I mean, they— in, in reflecting and thinking of them again more recently, a good friend of mine's brother was, was in Loose Shoes. Um, I was a bit of a groupie really, went to lots of their gigs. But maybe you were in the room at the time, I don't know. Um, but they're a large ensemble, and in a way, they represent to me that sonic version of allowing chaos at the edges, but with a strong sense of direction. You know, I thought their material was fantastic, particularly live. You know, you could argue not so strong in the records. So if anyone's actually picking up the recordings, then pick a live recording.

Okay.

Um, and then one more is Evan Parker.

Oh yeah.

Do you know Evan? Not well. No. Free improvising. He's at the other extreme. Free improvising saxophonist. I mean, that's probably a hard listen for most people, but amazing technique.

So kind of the Archie Shepp, Sun Ra end of things?

I don't know, it was like very long circular breathing. He looks like he's playing his saxophone forever. I mean, that's what I want to say for 2 minutes.

Yeah, I've heard of people who clearly practice circular breathing.

Hopefully I'm not one of them.

No, you're not one of them, Noz. You've been really interesting. Thank you very much for today.

Can I add one more thing?

Oh, go on then.

Because you're, just in terms of music recommendations, but then there's books. So I think there's a couple of books that I'd like to recommend to the group here because—

What do you think this is? Ezra Klein? Yes, go ahead, go ahead.

But since we're talking about formats, is it Between the Covers? Is that the other one on the television? Television, it sounds old-fashioned. Anyway, books. If you're in L&D or want to understand what's going on in the world in terms of the application of AI and technology, you can look back to some good fiction. And one of them is The Diamond Age by Neal Stephenson. I've recommended this regularly. It's a very good read. It's a prophetic description of how a young girl learns using technology. Really strong, great story anyway. Another one, Permutation City by Greg Egan. This is about digital twins and we're going to be, you know, creating a digital self. What does that mean? What are the implications? We're going to have to confront that very, very soon. And then one more, which is a very sort of practical one for if you're operating in any organization, is called The Art of Less by Alvison Spicer. He's quoted as saying, this is very relevant to what we're talking about, training is corporate sludge that doesn't achieve its goals. So the rest of the book is actually about, you know, look at systems, look at workflow, design about around support and improvement. Feedback loops is what it's about. So The Art of Less. Which is, yeah, I'd end on that. I think less is more.

Okay. And, um, on the subject of digital twins, shout out to the sponsor Synthesia. I think.

Ah, excellent.

Thank you very much, Lars.

Thank you very much.

That's all on the Learning Hack Podcast for this time. Many thanks to our guests and to our sponsors. The Learning Hack is completely independent. Independent and transparently funded by sponsorship. It is produced by Jon Helmer and hosted by him and also me, Mariana, a Synthesia AI avatar. If you want to help others find us, please like, follow, rate, review, and subscribe wherever you get your podcasts or on YouTube.

We'll be back in your feed in 2 weeks' time.

Meanwhile, stay curious, learning people.

Wow, now I finally get it. The Learning Hack Podcast is supported by Learning News, the learning sector's newswire. Rob and his team are good friends of the podcast, and we really value the help and advice we've had from them, and they do a great job. For the very latest news from around the learning sector, for interviews with learning leaders, the latest from learning sector vendors, and features on workplace learning, go to learningnews.com.