LH133 The Great AI Steal with Ross Stevenson

What are you willing to give up? How much of your soul are you willing to sell in order to get, in order to get some of this stuff where people are talking about, I've done this and it automates that? I'm like, yes, you may have done, but how much of your life did you give up to, to allow that to happen?

Welcome to The Learning Hack, a podcast about the people and technologies that are creating the future of learning. I'm John Howell.

Now guess what? Learning is cool. Learning is cool. Learning is cool. Learning is cool.

I'm learning.

Learning is fun. Knowledge is power. Knowledge.

Education.

Learning.

Welcome to The Learning Hack.

This season, we're featuring experts in Europe and the US who are helping learning professionals get to grips with AI.

Not theoretically, as in, will it kill us all, or where's my UBI, but hands-on in their everyday work. A couple of episodes ago, we talked to Erica Farmer. Very popular in L&D, very positive, encouraging, a friendly voice. This time, my guest is someone a bit more outspoken, a bit more hard-edged.

Hack Facts.

Ross Stevenson is founder of Steal These Thoughts, which began as a side project about learning and managing the monkey mind, and is now a newsletter, a think tank, and a training business. Before he went independent, he was Head of Learning at Trainline, and before that, he spent 6 years at Tesco running digital learning. Before either of those, he spent 7 years in HR. He's been at this for nearly 20 years, and he was training to be a software engineer when he landed in the people function almost by accident. But it's Ross's background in HR that influences his approach more than his tech chops. He says working in HR showed him how L&D looks from the outside.

It was like a 360-degree view of how businesses work because HR, for a want of a better terminology, is kind of like the police. They're kind of called upon when things are going wrong somewhere. You have to turn up and investigate and see all this kind of stuff, whether that's technology perspective, whether that's a people perspective or cultural perspective. So you get a really good understanding of an organisation's kind of culture and the people that are in it.

He describes his job now as AI enablement architect. And he expects it to become a distinct function in organizations within 5 years. Enablement is the key word there, and it's not just adoption. Adoption is licenses, logins, minutes on the tool, leaderboards, an AI week in May that nobody ever mentions again.

Most organizations want a one-and-done. You know, you come in, you do a couple of sessions, and bam, we're going to transform our organization. Everyone's AI native. We're rolling to the hills. We're making millions of dollars. And it's all a beautiful picture, right? We all know that's not the reality.

Enablement, something slower.

How do you work with these tools? How do you protect your own human cognition? How you think, the things that you're working on, how do you delegate to AI, and how do you work in a partnership?

Mapping how people actually do their work, finding where it snags and asking where the machine helps and where it doesn't. And if that sounds a little bit like performance consulting, well, he wouldn't disagree with you. So given that L&D has been promising performance consulting since forever, but not really delivering it, is AI the thing that's finally going to make it a reality?

Probably not. I don't think it's about the technology. I think a lot of this is not a technology issue. It's a people issue.

It's a people issue. It's about enabling people to use this new fast-developing technology, not about enabling the technology. It's an important distinction. Currently, Ross reckons 70 to 80% of the teams he meets have got no further than putting questions in a chatbot, and yet it looks from some research that interest in AI within organisations has already peaked. These are people issues. We talked about these issues and about what Ross personally will not hand over to a machine and why, and we asked another hard-edged question. What will be left of the learning function in 2030 if it continues to base itself in the content business? Here's Ross Stevenson.

So, hi, Ross Stevenson. Great to have you on. Welcome to The Learning Hack.

Thank you for having me. Privilege, how they say it. Longtime listener, first-time caller, all that good stuff. People will know you from Steal These Thoughts.

as I do, but there's a whole career behind that that actually, when I looked into, I've had quite unexpected. You came in through HR, not learning. Xerox, then Agilon, is that?

Or Agilon?

7 years in the people function before you had learning in your job title. What did HR teach you that you might not have learned had you come up strictly through L&D?

Yeah, very good question. It's making my brain cells think back to the younger years of my life when I'm sitting through these organizations in that function there. I think the main thing that I didn't get, I suppose, dropping right into L&D was that HR has such a breadth across the business in terms of where you're working with kind of different senior stakeholders. You're getting a lay of the land with different teams. And I think what it actually gave me, and this probably sounds bad for L&D, is an understanding of where L&D sit in businesses and how L&D is treated and how they're viewed. And I think unfortunately at that time for me, of those organizations was that it wasn't viewed very kindly and there wasn't really a lot coming from those L&D teams that was seen as value. And that stayed with me. And I suppose we'll get onto that as I made the transition into L&D myself around how that was actually viewed. And as someone in HR at that point, spending a lot of time with senior leadership and different stakeholders was really getting that understanding of how they viewed it. And also the other side of that is that the teams that I was in, they had very little L&D presence. So actually HR people were doing some of the L&D work, which seems a bit different nowadays. But back at that time, this was kind of like, you know, 2007, 2008, it wasn't uncommon where HR teams were kind of just doing a bit of everything. It was recruitment, it was HR tech, because I was looking after a CMS and an HRMS system, plus doing the kind of usual HR operation components on the side. So it taught me a lot more about that. And I say this to people quite often, it was like a 360-degree view. of how businesses work. Because HR, for a want of a better terminology, is kind of like the police. They're kind of called upon when things are going wrong somewhere. You have to turn up and investigate and see all this kind of stuff, whether that's technology perspective, whether that's a people perspective or cultural perspective. So you get a really good understanding of an organization's kind of culture and the people that are in it and the drivers in that. And when I made my transition to L&D later on, I kind of felt like that set me up quite well because I had a good idea of kind of how businesses work, the people there, and also the view of L&D in general. But like most people, I didn't plan to get into HR. HR wasn't my thing. I was actually training to be a software engineer. I just somehow happened to land into the HR space one day. And the kind of story wrote itself from there, from that perspective.

So Then you get into L&D, Tesco, then 4 years as head of learning at Trainline. Those are 2 very different organizations. One of them, an enormous distributed frontline workforce. And then on the other side, a tech company, Trainline, shipping software.

Yeah.

What worked in one and not in the other, and which of them shaped the most? How you think?

Both did. But that's also because, you know, they were really different kind of age ranges. I joined Tesco when I was like 25, you know, spent a great 6 years there. I just, they're a huge organization. At the time I was with them, it was 350,000 people that worked across that organization on an international level. So, you know, the breadth and depth was just something I didn't know and wasn't prepared for. Like, if you are from the UK, you think of Tesco, you think of the stores. Actually, I was working in was fundamentally the backend head office area. So I sat in a number of campuses. So the campus I sat in, which was just outside of London, had about 10,000 people. There were other campuses across the UK that had quite a few thousand more. There are campuses internationally as well. So I was working with predominantly all of that. I suppose we can call it backend because the frontline is what people know, but I was working with that population. You know, for me, that was kind of like, that is where, shall I say, I kind of learned my bread and butter and understand big businesses and how they work and the problems they have and the challenges that they have on the international playing field. So that shaped a lot of my thinking and how I understood the business world, how things work, different cultures, how things connect, because we were working on gigantic problems that took a lot longer to solve than you would do with some smaller businesses. Whereas Trainline, what it taught me differently was that, and there was a purpose, well, there was a purpose for making that move really, was that I wanted to go, at that point they were a startup, so they hadn't IPO'd yet. They were about to IPO in 8, 9 months' time. Never had an L&D function. There was a blank slate. No one had done anything before. They were looking to kind of get their house in order, um, before the IPO and go into a PLC. So I'd gone from Tesco where incredibly established, Lots of ways of working, incredible teams, really smart people, and soaking that all up for 6 years and then saying, if I was going to design my own L&D function, how would I do that? And then go into a tech company that was about to IPO to actually test those things and all those lessons that I'd learned from the 6 years at a huge organization. And like, what can I take across to that kind of tech organization? Now, the biggest differences that most people will probably figure out is that when I was at Trainline, there was the ability to move faster and there was immediate impact. So I could see the things happening in real time and coming back to that because the workforce had gone from 350,000 to about 800 people. So I was able to get away with a lot more stuff. There's a lot more experimentation. There's a lot more of a playground. Whereas at Tesco, that wasn't possible. It's not a bad thing. It's just that, you know, there's ways of working. There's a company that's been around for 100+ years. There's an establishment that you need to follow from there. So from a how's it shaped my thinking perspective, you know, because it's twofold to it. There's the big business of how do you work there at that level? How do you communicate to people? How do you make things happen? How do you understand technology and problems? And then from at that point, a startup IPO to go to PLC perspective is that how does that fast-paced world work that everyone wants to be involved in, especially in the kind of the London tech scene? And then how do you move from that in an IPO perspective to a PLC perspective? Like, and they're two like I say, 2 completely different challenges, but I think that's the beauty of them because they give you a lot of education in terms of, yes, I understand big corporates, but then how do all of the fancy tech startups work? So there was a lot there in terms of what works in the other. I mean, again, it's kind of night and day in terms of if you're at a company the size of Tesco, things take time. You're working on projects for years until you see any kind of ROI. And that's the reality of life. You're not going to, do something and then everyone's going to come on board and you're going to see the end goal in, you know, 6 weeks' time. That is going to be years in the making. But for something when I was like a train line, it was very easy for me to, and some people who are listening will be familiar with this kind of onboard suppliers, build learning technology, bring all that stuff in rather quickly and test it and bring it out to an audience and get that feedback and then able to iterate on it. So yeah, they both have their place. I always say to people who are asking about careers is, I know there's a big kind of like movement, definitely in the last few years, where people want to go and build their own thing and do their own thing out of the gate. But I'm always a big proponent of go and work for big companies, go and understand how business works. How do these challenges work at those organizations? Because I truly think that will help you going forward and doing whatever you do, because you've got that kind of experience. And something I've spoken about with David James in quite a bit of detail is you've got that kind of career credibility. You've got the understanding of You've been there, you've worn the hats, you've got the t-shirts, you've got the scars, and you can actually go off and connect to those people in the real world today and what they're trying to do.

So in 2023, you go to Filtered, a company we know well here. We've had a couple of guests from there, Toby and Adam, who's now at AssembleU. It was a significant move because it's kind of going from being a poacher So it filtered as a vendor selling AI into L&D. You went there as a learning strategist, weren't there for all that long. And then you go independent. So you've gone from one of the— I mean, Tesco is the— people in the US might not be aware of this. Tesco is the largest retailer, I think, in the UK. They are, yeah. You go from a quite massive company down to a smaller company, down to essentially a startup. Or scale up Filtered and now you're an independent and everything since then has been about AI enablement. So with the Filtered move, you crossed over, as I say, from somebody who buys the stuff to somebody who sells it. What prompted that move from practitioner to vendor?

Yeah, it's a great question. I mean, and being completely transparent, it wasn't in the plan. So I was always like, never going to work for a vendor, not going to be part of this because obviously as being a practitioner, you know, we've all been to Learning Tech, we've all been to the other conferences where you've got vendors left, right, and center sending you all this stuff. I'm most certainly not a salesperson, so that was not my field of expertise whatsoever. But the filtered thing was actually because I knew a lot of the team over the years through building Steel These Thoughts, through writing the newsletter, talking about stuff. And actually, and this is one of the great things about doing the things that I do is I've been able to connect with a lot of great people. And I was able to connect with people like Mark Sale-Sanders, Toby Harris over a number of years. And we developed a relationship. And, you know, I come to a point in my career where I had no interest continuing in going to director of learning, or going up that ladder at all. I was kind of like, I would love to do something different. And those paths kind of just converged at that point. I was speaking to Mark and Toby at Filtered. I was really interested in what they were doing with content intelligence, the companies they were working with, some of the problems they were trying to solve. And I love smart people working on interesting problems. So it was just one of those opportunities where it was like, here's the door, my friend. Would you be interested? And I was like, yeah, let's do it. Let's see what life is like on the other side and how can you contribute to that. And I think the great thing there was that was really understanding because I'd been a customer and then been able to speak to all of those customers and see what life is like on that end of the table, if that makes sense. And then go through that experience with them. I kind of understood that mindset already of what they were looking at. And also to look at the inside of an organization that was really one of the first for a long time selling AI solutions before it became sexy into businesses themselves.

Yeah.

So, and I must credit that team as well because actually that team was the team that introduced me to early-stage LLMs. So before there was ChatGPT, the team there was already talking about GPT-2.5, GPT-3. If you're a super nerd, you'll know those acronyms. If you're not, you'll need to go and Google them. But they were the pre, I suppose, bits of tools you could use before ChatGPT, where it was a playground interface. So yeah, a lot of that kind of more understanding of what we know as generative AI today actually was born out of that environment.

Oh, interesting. So Still These Thoughts started life as a side project about learning, personal development, and managing the monkey mind, and really nothing to do with AI at all, although obviously you knew about it at the time. 7 or 8 years later, it's a think tank, a newsletter, a training business, and a job title. When did the AI thing take over, and did you choose it or did it choose you?

Yeah, so the AI thing wasn't chosen. So the context is I've always been in learning tech and HR tech. So regardless of what my role titles have been, I've always been involved with tech in the HR and L&D world. And AI was just a natural evolution and kind of going back to the filtered component. Once I'd seen GPT-3, which was pre-ChatGPT, I was like, oh, this is interesting. And I was experimenting a lot with that. And then when ChatGPT came along, there was this classic tech terminology, the first mover's advantage. I was already really familiar with what these companies were trying to do. I already played around with that stuff. I could see the applications. And then when ChatGPT landed, I was like, okay, this is quite cool. Let me play around with that and see what happens there. And like most, I suppose, nerds and stuff, I started building in public. I started sharing videos. I started sharing loads of different kinds of resources around the things that I was doing just for experimentation. There wasn't really any— Vision of I'm going to do something with this. I thought, oh, this is really interesting. This seems like something that may be worth sharing with people. So I kind of just did that. And there's still these thoughts. Newsletter at the time was already talking about for a long time, learning technology and ecosystems and how you build your stack and all of that. And naturally things just changed. People stopped asking me about LMSs and LXPs. They were more interested in, oh, I saw you Did this with ChatGPT. How did you do that? Can you tell me about ChatGPT? What is this AI thing? What is generative AI? So it kind of just organically moved in that direction where, as I say, people didn't want to talk to me anymore about the traditional things that we were talking about because everything had moved to AI. So you could really say they chose me and there was part of a kind of adapting in there and grown with me as the more I was sharing of this stuff, the more people wanted to talk about it and go down that avenue. So that's, yeah, that's really how that all came around. I never set out to have really anything to do with AI. And little did I know back in the days of GPT-3 and playing around in the playground that it would lead me to this environment where I am now, where for some reason people all want to talk to me about AI. I'd love to talk about more than AI with people, but obviously AI is the sexy trendy topic. So that is what it is.

Yeah. You've got a good background there to talk about it. in practice and on vendor side. So it's a useful pivot. And that brings us up to date. Some people might be astonished we've got this far into the interview before I really got into AI seriously and started asking you the questions that people ask you all the time now. So you argue that L&D has to move from AI adoption to AI enablement.

Mm-hmm.

Counting licenses and logins is measuring the wrong thing. just kind of activity measures, not outcome. But can you unpack that for us, the difference between adoption and enablement? What does an organization that's genuinely enabling look like as opposed to one that's just deployed AI?

I think the central thing with enablement are people that are actually treating it as part of the organizational culture. So what we're not talking about here is these one-off events where And we all see them, right? There's AI weeks or AI months, and then we kind of never focus on this stuff again. It's actually looking at saying, how does this become part of an evergreen strategy that's embedded into the things that we do? And a lot of that, which is a popular term at the moment, is evidently workflows and workflow mapping, which is basically just saying, how do you do what you do as an individual? How do you deconstruct that and then understand where may AI help within that process? And that's more focused around enabling, which is actually saying, thinking about how we work, how we deliver stuff, and how these tools can become part of that. And of course, it's always changing because the more these AI companies release new features, the more capabilities are available for you to then say, oh, okay, I couldn't do this 6 months ago, but this is possible now.

Yeah. Isn't that slightly— sorry to interrupt— isn't that slightly maddening? I mean, I found just this morning, I was kind of going through my workflow. Again, I thought this is about the 3rd or 4th time I've done this this year. Because every time you reorganize your file structure and then suddenly, well, you don't need to do it on your machine, you can do it in the cloud or whatever. Or suddenly, I can drive Claude from my mobile phone, which I've been waiting to happen for a while.

Yeah.

But things come up and a load of work that you've done, previously to change your workflow suddenly gets junked because the things got smarter, the things moved on. As you say, an event isn't good enough. It has to be a constant activity.

100%. And there's a great example that I've had with a lot of companies over the last 6 to 12 months where at one point it was a big deal where you could create custom AI assistants. So Gemini had Gems, ChatGPT had GPTs, You know, at that point, it was kind of the frontier. But now we've had the evolution of skills and markdown files, which have fundamentally been the evolution of that. And again, that has then needed companies, needed individuals to essentially reevaluate. And as you're saying, kind of rewrite what they're doing. So you are always kept on your toes. And I think that's the beauty and curse of enablement is it never ends. Because there'll always be something new that comes down the barrel to then say, okay, in light of what we now have today, what can we achieve? What changes? So enablement, like I say, for me is really focused around that. Whereas adoption, as I've kind of spoken about, and probably people think I'm a broken record by now, is very much focused on the, from an L&D perspective as well, the typical perspective of how many tools do we have, the licenses, how many logins, how many minutes or hours are people spending on tools versus versus actually saying, how do we derive value from those tools long-term rather than just a kind of like vanity metric, shall we call it?

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.

Do you think people who manage L&D have really got the change yet in the sense of how different it is to work with AI?

Easy answer, no. And I think I see that quite often with just through discovery calls of organizations. So even if I don't work with some of those organizations, just the general discovery calls and Just listening to people for 20 to 30 minutes to get their understanding of what they get about these technologies today, how they're operating with those technologies. I think the majority is, yeah, it's a no. They don't really understand what's there. The typical scenario is there is a mandate in those organizations from senior leaders, which is saying we need to use this stuff. And then you, As L&D teams need to make that happen. You need to make that stick. And it's kind of like this Wild West really in terms of, well, what does that mean? How do we do that? How do we do this thing that we don't really understand?

So you've named the job you now do as the AI enablement architect, and you reckon that's going to become a distinct function within 5 years in organizations. And when you kind of list what the job entails, diagnose maturity, map to the workflow, design the rollout, embed support, measure outcomes, not completion. and so on. It does occur to me that my good friend Guy Wallace, or going back further, Gilbert Rummler, would recognize every line of that. It looks a lot like performance consulting, doesn't it? Nothing wrong with that.

It is.

It's all very sound thinking. But the problem is L&D's been trying to do that. It's been promising that actually since before either of us even started in learning, certainly before we started podcasting and blogging. Why should it work this time? Is AI really the burning platform that's going to make it happen at last?

Yeah, probably not. I don't think it's about the technology. I think when we look at these things, and obviously there's loads of narrative online about this as well, that a lot of this is not a technology issue, it's a people issue. It is a concept of helping people to move forward and understand and leverage these bits. And as you said, like, AI enablement architect, as I explain it at the moment, is for a want of a better terminology, because I'm not really sure what to call it. It's basically just the services that I'm asked for from organizations and the data that I get coming through. But it's most certainly just performance consulting. It is solving problems. It is layering that onto the technologies that we are using today. And the honest answer is, yeah, I don't know if it will stick. And I don't know If it will be any form of revolution. A lot of that comes down to many components industry-wide. I think people in pockets can make a lot of change, can create a lot of transformation based upon the culture of their organization, the context in their organization. But industry-wide, and look, we all know HR and L&D industries are very slow to move on things. I'm, Yeah, I'm not so sure at the moment in time, but I like to be an individual much like Guy and others that talk about these things, that think deeply about these things and look at how could we make that happen? What would the enormity of the possible be if we can imagine those things and introduce those things? So yeah, I try to place my hat there, should I say.

Yeah. So I'd like to talk a bit more about looking forward to the future later, but let's get a bit more detailed about some of your engagements. You've written up Galen Crawford at Stripe and Chime.

Yeah.

Canva's AI Week. And the Canva detail that was interesting there wasn't the training, it was what came after it. AI hubs, fortnightly forums, exemplar networks. Could you walk us through the process of working with an organization from cold start? I mean, you mentioned a little bit about discovery calls there. How does it pan out from there onwards for you? I mean, obviously you're probably NDA'd up to the hilt, but—

I am.

If you could give us some indication, as am I, what do those engagements look like week to week?

Yeah, it turns into a media interview here where I carefully select my words and bits that come through. So your audience will have to forgive me as I might just go silent for about a minute as I think through and pull out those bits there. It's all right.

I'll edit the silences out.

There we go. So I think the top level is, look, you know, how most engagements go is there is this discovery session. And I think the discovery session is one of the greatest things to have because you really get an insight into, as I was saying previously, what do people know about the technology? How do they think about the technology? How do they think about the way they're working with the technology, both themselves and in their organization? And really importantly, What are they actually trying to achieve? So I think that performance consulting bit that we spoke about comes nicely into this because 99% of the time, a lot of people that I engage with don't think about that. They, for whatever reason, they're not thinking about what do we want to achieve? It's just a, we've been told we need to make this work and can you come help us make this work? So not really looking at the actual problems to solve there. Once we go through that discovery session bit, essentially what happens is it can go off in a number of of different directions. We can move into looking at getting deep and down into the details on a local level with specific teams. Maybe it's departmental level, maybe it's organizational level, which is a different beast in itself. But for me, it's then really starting to understand what are the friction points that exist in the organization today at those different levels, and where could we possibly help And then with that, the layer of AI in that, where could we use that in there potentially as part of that to help with some of those problems, to help with some of those frictions? Because I always think about what you're looking for in these scenarios is where is the pain? I don't necessarily care working with organizations about how do we use ChatGPT or Claude to make our email sound more sophisticated or to write this one-page document. It's actually more interesting, especially for these businesses, to look at the problems to understand, well, Where are the breaks in their process currently in their L&D team as an individual or as a team or as a department? Where could AI in its many guises from a generative AI perspective support in that, which goes beyond the casual stuff that we know? So they go into that sphere. And like I say, after that point, it can go in many, many different directions depending upon what are the things that we are actually trying to achieve. And then And the levers that are pulled will vary across that from a week-to-week basis. But the ultimate goal is always to make it part of the organizational conversation and context. It is not to be, we work together for weeks or months or maybe a year, and this is treated as we just do this then, and then we don't do anything for the next 10 years of our life. It's very much about how are we building that system to then say that you will continue to do this stuff, continue to evaluate workflows based upon the latest technology, the latest way of thinking, all the things that we're doing, and actually drive that for value in your organization. So at a very high level, without getting myself into any legal trouble, that is basically what it looks like.

All right. And how would you do that specifically to keep it going over time, to keep it sustainable. And do you ever get any pushback on this as a consultant? Something I sometimes worry about is that I look at going into an organization advising them more often about how to start podcasting, use that for learning and so on. Well, the way to do this is really to embed it, to talk about your brand and also to do it long-term and to review and so on. So really you need me to come back in in a couple of months' time and a bit after that, and maybe do a kind of weekly call and all the rest of it. And a client can look at that and think, well, you're just trying to write yourself some extra work.

Yeah.

Do people push back on that? And where is the tension in what you're doing between this needs to be long-term and sustainable, and I don't want to look like I'm just kind of Trying to rack up consultancy hours here.

Yeah, 100%. Look, there's pushback all the time. And that is mainly because, and I'm sure this won't be too unfamiliar to listeners, is that most organizations want a one and done. They want, you know, you come in, you do a couple of sessions and bam, we're going to transform our organization. Everyone's AI native. We're rolling to the hills. We're making millions of dollars and it's all a beautiful picture, right? We all know that's not the reality. But most of the time you have to work and going back to the kind of performance consulting methodology here, you're working with these people to help them understand that. Does everyone want to do that? No. There are many cases where people just say, look, at the end of the day, we want a few workshops on Claude, or we want to work a few workshops on amplifying what we're doing with AI. And that's it. And sometimes that's the gig. And it's up to you as a consultant, whether do you want to do that kind of work or do you not want to do that kind of work? The ones who are more interested, of course, are the ones who want to do that long-term, who want to spend a year working with you doing that. But to your point, it is very hard to show them what that payoff is going to be because there's no proven blueprint, right? What we can say is look at past technology implementations, look at the things we do at L&D previously, is that behavior change takes time. None of these things are going to be an overnight success. You know, there's very many examples of everyone's talking about the LMS and NXP at the moment and how those things are changing. They will change. But it won't change in the next 4 days. It might change in the next 2 or 3 years and how we look at that. And much like a lot of the work when I was at Tesco, I was responsible for digital learning, the learning platforms there. And a lot of things we were trying to achieve, they took years. It was 18 months to 2 years until we actually saw the huge benefits we were looking for because it took time, as you were saying, podcasting is there. We had to build engagement strategies.

Yeah.

We had to be consistently focusing on how are we keeping people up to date with advances? How are we positioning people? How are we helping them transform? How are we showing them that transformation? And this is the same kind of challenge we have now with AI is that it's not saying, here's a ChatGPT license, here's a 1-hour ChatGPT session. It's actually spending months with people to go through, like I said, the work that they're doing, the challenges that they're facing. How can those tools support that? Coming back to those people to then look what are the behavior changes? What are the habits? And how can we continue to support those and bring those to life? But, you know, as I say, it's few and far between. You get many more organizations that just want you to come in and go, come and do a few workshops. We'll hit our arbitrary goal, whatever that is, 90 or 95% AI-native marker in our group. And that's it. But a lot of those companies won't I'd say succeed long-term.

Yeah.

Because they're not really focusing on that, you know, to the podcast and analogy there, they're not really focusing on how does this become part of what we do daily, weekly. They're kind of looking at it as one-off, this is it, and then you are left to your own devices. But as many practitioners will know, leaving people to their own devices to just figure it out doesn't end well.

At the really simplest level, in a way, how are you finding people's abilities with using AI and what are you doing to move them onto the next step? I mean, are you helping them to build more context in the way they structure their use of AI? People talk about context engineering now. I know Adam Lacey's got a webinar about that. You have to kind of build up the right type of files for the LLM to consult before it can do useful stuff for you. You know what the devil's bargain is? The more it knows about you, the more helpful it can be.

100%.

But you have kind of privacy issues. And so are you helping them through that? Are you taking them on to doing something more agentic? Where are they and what are you doing to move them forward?

So, if we have the context of most of the teams that I've worked with, and I'd also say many people in the news there as well, a lot of it is really kind of helping them understand what do they have in their palm of their hand right now. And I don't think I'd be wrong in saying maybe 70 to 80% of the people that I encounter are predominantly in what I call the AI chat field, which is basically, I will go to Claude or my LLM of choice to ask very basic questions, very basic information retrieval. And that's pretty much it for most people. And that's really because they haven't looked beyond that. No one's actually shown them to ever say, well, here's what else you could do. You can start using MCPs to connect into third-party services that then give you access to more data that then help you Kind of build narratives that help you do other things that weren't possible before. Other elements of tools that help you maybe go into a design field where you can extend your capabilities without having the skills to do that. That's kind of where I'm bringing people currently, because I still see a lot of teams are maybe 1 to 2 years still behind on where the current tech is at, but that's completely fine because it's literally impossible to keep up with every single new thing that exists in the world today. And how to leverage that. So it's taken people from the maturity of everyone's using AI. That's kind of a fact right now. Most people understand AI chat. What they don't understand, and yes, you know, kind of context engineering, or I kind of call it context layering, is really the next bit for those people where it's saying, well, how do you get more from the tools? So how do you start using skills in ChatGPT and Claude to do more work for you? But in that, you need to understand how you do that work already. How do you add in other services? So I was talking on LinkedIn the other day about the concept of headless LMSs and headless LXPs, where it's basically just saying, instead of using the LMS's UI as the front door to go in and use the services, you can connect that into a Claude and ChatGPT, and you can then engage with the content libraries, you can engage with the features through that. So it's bringing people to that level. And I would say right now, it's, I don't think a lot of teams are even in the sphere of understanding agentic AI, true agentic AI, not what the product and marketing teams are calling, now calling AI agents, what they were calling AI assistants a couple of years ago. Of course, there are pockets of people at larger organizations that are further ahead for sure. But for the majority, it is that graduation from what we call AI chat to How are we bringing on board all of these other services and these capabilities that they have access to? They just don't know how to use them. And quite frankly, not really sure why they should be using them in the sphere of their world currently.

Yeah, interesting. So it's kind of about context and connections is where it's at at the moment.

100%. And you could do a lot with that at the end of that. I think people get into that level, they would open up a whole new world for themselves than what they are currently doing. I also think it's just a little bit of education and understanding. I kind of call it your workspace. So how do you build your AI workspace? How do you, and again, there's a lot of data and privacy stuff around here in terms of how much you want to give away.

Data is such a hot topic.

100%. So, you know, from an organizational context, from an individual context, you need to think about those caveats in there. But when you're creating that AI workspace is the that these tools will give you more depending upon how much you are willing to give up to it. And I think that's an important kind of balance that we all need to consider in that is that some of the things that you see online are achievable. If they're achievable in a corporate context is a very different question.

Yes.

And also, you know, what are you willing to give up? How much of your soul are you willing to sell in order to get some of this stuff where people are talking about, I've done this and it automates that. Yes, you may have done, but how much of your life did you give up to allow that to happen?

Yeah. So there's also a question of, it's something we talked about before I hit record, which is the kind of personal discipline of, to what extent can I let it draft my emails? And to what extent do I want to let it make it sound a bit more like me and actually take over and draft the things myself because I don't want to get accused of work sloppiness.

Yeah.

Which is the thing and so on. So that personal discipline thing we mentioned, like I've got a deadline. So you kind of let your morality slip and you let it drive something for you that really you should take for yourself. Those kind of personal disciplines I think are things that have to come into the L&D canon in a way, in the same way that people talk about things like resilience and time management.

100%.

Yeah. What not to do with AI, I think is actually a way more important question than what you can do with it.

Yeah.

And there's loads, you know, there's loads of quotes that I can give off. And I love throwing out the old Jeff Goldblum ones from Jurassic Park about the scientist one, you know, your scientists didn't think about just because they could, whether they should. And all of these kinds of things. I think that more of my conversations nowadays have gone down that field where I'm saying to people, you know, there is a huge menu. of what you can do. But actually, you should really consider what do you want to protect? What do you not want to outsource? And what shouldn't you outsource at the end of the day? And as we're discussing here, you know, for me, it's always been my writing. Anyone that reads my stuff, I would, you know, I would describe some of my writing as probably chaotic, unhinged, full of sarcasm, but it's very hard for AI to copy me at the moment in time.

That's why I like it, Ross.

It's madness. It is honestly, it's how my mind works. But there's a, you know, there's a distinct, if you saw something that wasn't that, you'd be like, that's AI. Because that's not how he talks and projects himself. So I truly do think there is a, for everybody, organizations, individuals, there is a component of what do we not allow this into? Because what do we enjoy doing? But also from a skills perspective, from a cognition perspective, what are components that really help us as humans? And as we were talking about before we hit record is, and I agree with you, Writing being thinking, and I do my best thinking when I write. And whether that's through a keyboard, whether that's through, you know, I'm using an app called WhisperFlow at the moment to basically, there we go, to vocalize all my thoughts. And I'm finding that an interesting experience in itself to do that and think out loud. All of my best stuff comes from there. And people will think I'm crazy, but AI has yet to defeat me on me doing my own writing and doing my own thinking out loud. So for me, that is something I will never outsource and will not allow AI to do for me. It could do research, you can do analysis, but the actual component of the writing, constructing those thoughts, I'm a weirdo as well. I get a lot of like kind of like pleasure out of doing that. And oh God, that's really, you know, I loved how that was written or that analogy worked there really well. Or it was a great sarcastic comment and a meme or a GIF that would work really well with this component here. I love doing that in my mind. I just don't want to hand that out, so I won't do it.

So here's an exemplar of just what we've been talking about. I'm going to give you the beginning of this question in the way that Claude wanted me to say it. Hey Ross, here's the Monday morning question. I mean, that's something I'd never say, but actually the Monday morning question is quite an interesting thing. So let's think about a head of learning might be listening to this on the way into a job where the AI strategy is a license and a lunch to learn, not much else. They've got no budget, they've got no mandate. What do they do this week that a year from now they'll be glad they did?

Yeah, my first thing would be definitely get a better Monday morning, that's for sure. My second and more serious answer would be this all depends. And I'm very careful with what I say here because I never want to get people sacked from their jobs and what they're doing by being kind of gung-ho. But, and the reason why I say that is because any advice that anyone gives you in our industry or other industries really depends on the context in your organization, the culture in your organization, and the constraints that you've got. Because I can say loads of stuff, like many people can, but it could be absolutely wasted, depends upon those 3 levels of context, culture, and constraints.

Yeah.

So, So my basic advice is always going to be, look, focus on solving the actual problems in your org with AI. So that for most people, it's not going to be, how do we write better emails? Or how do we produce trillions more pieces of content that are probably far more subpar than what we used to do 5, 6 years ago? So it's looking at our pockets of where is the pain? Where is that friction? What is the pain that you can provide. And that can be very, very simple things. There's basic stuff for people that we've just discussed. Maybe that could be helping people understand what more can they do with the LLM or other AI tools they've currently got in their organization that can directly impact their work in some way. I'm not going to use the words efficiency or time-saving because they are very kind of generic outcomes that people look at there. The biggest thing for me that comes to mind is, and some people may disagree, but I think for me is focusing on AI fluency. So when I say AI fluency, that is not just how to use tools. That is understanding how do you work with these tools? How do you protect your own human cognition? How you think, the things that you're working on, how do you delegate to AI and how do you work in a partnership? That is incredibly important to me. I think if you spend any of your time as an individual or as a company in the next year, it is focusing on that AI fluency piece and understanding not just the tools, but how do we do all of this other stuff? How do we work with those tools? That will set you up, in my opinion, far better as a team, as an organization, and as an individual, instead of just saying, I can use ChatGPT. But if you can understand how to think for yourself, have that independent thinking, amplify that with ChatGPT, use it to actually execute on the things that you want to do, that's way worth your time.

Let's pull focus a bit. I kicked off this year by talking to Don Taylor about his Global Sentiment Survey. Always interesting to talk to Don, which says interest in AI has peaked and that people report AI is making their stress worse, not better. You've said something adjacent yourself. Most teams are still using innovations that are 2 years old and hype has become the enemy of adoption. So are you selling enablement into a rising market or a receding one as far as interest goes? I'm not talking about stock prices because obviously everything goes up and to the right with that end of AI. I'm just talking about people's interests. They're kind of fed up of hearing about it. And if we have peaked, it feels odd to be talking about interest in AI peaking because at a time like this, is the peak the problem or is it the opportunity?

Yeah, great question. I think it's most certainly the opportunity. And I think with the enablement piece, for sure, because adoption, as we kind of categorized earlier on, I think, and I see it on my own newsletter, right? I'm always doing kind of surveys of people in the news there. And it's very clear, let's say before this time last year, there was a huge amount of AI anxiety. But what I see now from people in that data is actually, well, how do I make this work? How does this work operationally? Like I've seen a billion tools now. I don't need any more tools. This is the tools that we've got. How do we actually make this stuff work and make it give value? So I think there's huge opportunity there. In the enablement space for people to come in. Some of the words and phrases I've been using already is to actually look at the friction, look at the pain points and say, right, how do we solve those pain points? What is the painkiller? And enablement in my kind of language here is the kind of painkiller to that. How do we help people with that and show them that? Because, you know, as Don says there, and I've seen too, is that, you know, people don't want to talk about tools anymore. They're sick of talking about tools. We know about the tools. What we want to know is now, how does that make me look good? How does it make me look good? How does it make me get a promotion? How am I going to be able to do all of these things? Because they're still the drivers that, you know, bring people around to learn. And if you say to people about, if you come to understand how this thing works, it's going to make you look great at work, it's going to potentially give you more money and better career opportunities, you're going to go down that avenue. And that's what people want to do now. They want to see the ROI. And I think that's the big question. And we see this a lot in mainstream media as well. well, is a lot of companies now asking the question, okay, we've had this stuff for about 3 years now. What's the impact? Where's the value from an impact perspective, and also from a monetary perspective? And, you know, I think there's a great opportunity for more people to come in and work on that side of workflow design, looking at the pain, looking at the friction, what can we learn there? And where could AI unlock some of that? It's not going to be the answer to everything. I think that's an important thing to say, because the as I coined them, these AI bros talking social media. You think AI is the solution to every single problem out there, and it most certainly is not. But there is an opportunity with enablement to unlock a lot of this stuff.

Right. So as the hype recedes, really, it's an opportunity because we can just kind of get down and use the stuff for what's useful.

100%.

Yeah. And the hype really doesn't help.

Okay.

So Looking forward a bit, let's really depress ourselves. If we look forward 5 years, L&D doesn't make the move you're describing, it stays in the content business while everyone else works out AI for themselves. What's left of the function in that scenario? Is there a version of this where L&D simply doesn't survive the transition to working with AI?

Yeah.

Are we going towards the end of the interview now? So it can't all be down. What about the more positive version? Where do you see green shoots and reasons to be cheerful? And maybe 5 years is too much. Most people talk about 3 now.

Yeah. Okay. So let's start off with the dystopian view. I mean, obviously I don't know. This sounds like I'm reading the tea leaves and it's kind of like a lot of dubious speculation here, but maybe it does stay the same way. Maybe all that happens is that people just keep building content, but they keep getting AI to do it and it goes down that avenue. I think the common sense outlook is, does that end well for L&D? Probably not. And that probably means a lot of reduction, a lot of transformation. Some of the signals that we are seeing already where there is opportunity for that script not to happen is if we look at organizations like IBM, like Atlassian, I think I'm saying that correctly, and Zapier as well, that have basically been merging their, some of their engineering teams and their people teams to become enablement teams. And the classic CHRO role or CLO role is now becoming a head of enablement across an organization, is that we're seeing some of those saying, actually, well, it doesn't have to be that dystopian view where it continues to be content houses and focusing in that way. These companies are signaling and kind of saying, there is a model here where we really want to make this work. And we understand it is a people and a technology problem to solve together. So I think there are those 2 routes. I don't know if you have read it yourself, but there was this kind of paper, fictional paper put out about, I think it was AI 2027, where they were basically pontificating on what could happen if the current trajectory continues the way that it is. And it gave audiences these kind of two views. It gave them that Orwellian dystopian view, and it gave them a view of, if we get this under control with governance and more governments coming in to look at this stuff, how can we leverage AI? And I think there's a similar bit for L&D in that perspective. L&D has that same route there. So it has that view of, you could continue down, let's classify that as that dystopian alley, which is probably just continuing to do things as they always are, but amplifying that with AI. Or we can see some of this route happening now. Like I say, those organizations where they're shown a route to actually look at, well, how could we transform? And what does that involve? And what does L&D take on at that point? And you could kind of set the names aside of L&D. Maybe that focus is more in the realm of pure kind of transformation, change, whatever kind of element that comes in. So I think there's a good model for that. The reality is it's going to be a choice for for individuals, for companies, and that will sway the industry to look at, you know, what will that be going forward. We're not going to remove all this stuff overnight, and you most certainly should not do that. But actually, I think it's going to be looking at, okay, well, there's some components here. What are we going to choose? Kind of Squid Games a little bit. Are you going to choose yes or no in terms of if you're going to continue kind of doing this round of what we're currently doing in L&D? So what I would say is kind of the TL;DR is on is I am very hopeful, but it's going to be based on individual and group choices on what the outcome will be, let's say by 2030.

Let's hope it's not Squid Game. That's quite terrifying. Okay. Always the last question for fun guests. Favorite karaoke standby?

Yeah. So it's going to be In the End by Linkin Park. I've probably heard that song about 100,000 times at this point in my life. So yeah, that'll be the one.

Which one is that?

In the End.

Oh, In the End. Sorry, I thought, yes, yeah, not, yeah, I think I can remember about one Linkin Park record, but you know, different generation. Okay, marvelous. Well, I'll see you in the karaoke suite.

That will do.

Thank you very much for that, Ross. I really enjoyed that. And I think people really have learned something out of that.

I hope so, and thanks for having me.

Thanks for being so generous with your time.

Cheers.

That's all for this time. Thanks to our guest and to our sponsors. We'll be back in your feet in two weeks' time. Meanwhile, stay classy, learning people.

Now I finally get it. Let me ask you: Have you learned anything from this podcast? And are you wondering maybe whether a podcast could help your people learn? Well, when I'm not making my own shows, I walk organisations through making podcasts for learning of their own in-house. And now I've written a paper setting out the evidence for where it works and where it doesn't, together with 10 questions that will help you decide. It's free to read or download, no form, no follow-up, at johnhelmerconsulting.com. 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 For the latest on selected vendors and features on workplace learning, go to learningnews.com.