Chris Van Wingerden: hey everybody. Welcome to Instructional Designers In Offices Drinking Coffee. #IDIODC. As always, Instructional Designers In Offices Drinking Coffee is brought to you by the team here at dominKnow, makers of dominKnow | ONE, helping learning and development teams develop, scale, and deliver learning that maximizes employee value. Check us out, dominknow.com. the website's there for you 24/7, even when we're not here. Gang, this week we have Donald Taylor back with us. Donald, you were with us last year. You're back with us this year, and the reason for that one-year interval is the annual global sentiment survey in our space that you conduct every year. and we had a great conversation last year. let's start off talking a little bit about the survey itself. Tell us, how you gather the info, the groups and folks that you're connecting with on this. Remind us, I guess, what the data points, or the data gathering, how that works.
Donald H Taylor: It's a shamelessly simple, short survey. I have like five questions. But I've been doing it for 14 years, and the idea is the more questions you have on a survey, the less likely it is that anybody completes it. So I keep this thing short with the idea that it goes wide. So we have, this year, 3,700 people from over 100 countries responding, and the key questions on the survey are, what do you think will be hot next year? and they have a list of 16 things to respond from. they can choose one, two, or three options from that list, and everyone, almost everyone chooses three. The other two are, what's your biggest challenge? And the last one, which is new this year, is what have you done that's new in the past 12 months? I'm gonna keep that same format going forward. But the key one, What do you think is gonna be hot in the year ahead? That's been going for 14 years now, so quite a lot of longitudinal data on that. the challenges one, what will be a challenge for you, has been for five years. That's pre the launch of ChatGPT. It's really interesting to see how the challenges have shifted over that time.
Chris Van Wingerden: and by your reference to ChatGPT, it sounds like your data had a lot of focus on, or the feedback that you got had a lot of focus on AI this time around, which, you know, totally, I think for Paul and I would be a total surprise. It's certainly the dominant conversation in everything, isn't it these days? I'm sure it was, like online learning and how can we move things online now?
Donald H Taylor: You know, I was looking back at the conference schedule for Learning Technologies. We had Learning Technologies last week in London, and I was looking back at the conference schedule for 2003, 'cause I've chaired this thing since 2000. And, I was looking back at 2003, and we had on the schedule how to move your learning online. Yes, we also had how to measure the impact of learning and, how to make people care about learning. And some things just don't change over time, but some things do. AI is absolutely shifting what we're doing in learning and development. And I think importantly it is removing one of the things that learning and development's always been able to say was its foundation, which is the creation of content. That's shifting now, and I think that's probably gonna be a key point of discussion here. I'm just seeing the chat. Uh, we got a lot of people saying, a lot of people coming from, from different parts of the world. It's great. Let's move on. But that's the point, AI is changing what's happening in L&D partly because of what it does, but partly also because anybody can now do it, and I think that's the crucial change.
Chris Van Wingerden: So, what kinds of things would have been identified, I guess, as one of the hot things. what kind of things were then being revealed? What other things were revealed though as hot things still?
Donald H Taylor: the question what's hot is a really stupid question to me because I don't define what hot is. And people can choose their own definition. But nonetheless, over the years, there's been very predictable, patterns of behavior in response to the question. So, what happens is things start off being very popular, and then they decline in popularity over time. So, mobile delivery was the number one choice 14 years ago, and I took it off the survey two years ago because it's now business as usual pretty much. Not everywhere in the world. other things go down not because they've become par for the course, but simply because people were excited about them, but they're not so excited about them now. Like curation, for example. Once very hot, meh. People discovered it was a bit more complicated than they thought, and their attention shifted elsewhere. Personalization has been doing exactly this trend. It was coming down for eight years, and then with the launch of ChatGPT, it started going up. And it's gone up for three years on the trot, which is completely unheard of as a pattern. Normally when things go down, they stay going down.
Paul Schneider: Mm-hmm.
Donald H Taylor: what's happened is the rise of artificial intelligence, which has also gone up for the last three years, that has led to other things riding on its coattails, but particularly personalization. So with the rise of AI, people are saying, "Ah, I can use that for personalization," so that's gone up as well. Strangely, you could also use AI for learning analytics, but that has fallen away for the past three years. So people are making a choice. They're going for the content piece, they're not going for the analysis piece. I think that's interesting. The other thing that's gone up for the past two years is showing value, and showing value normally doesn't go up or down. It stays around 6.5 to 5.5% of the vote. But it's broken out this year, and it's hit the highest point it's ever been on the survey for 10 years. So something's happened there. People think personalization's important and showing value is important, at the same time as, of course, thinking AI is important. It's worth noting that AI was number one for this year and last year and the year before. But this year it actually dropped down by a very small percentage, .1%, but it did drop down. So it's sort of hit the peak of its level of interest. Doesn't mean we all know what we're gonna do with AI, but it means that the height icon has peaked out. Now we're gonna work out what do we actually do with it, I think.
Chris Van Wingerden: Yeah. It's moving into a defacto kind of thing as opposed to an upcoming, start of the curve kind of a thing. Yeah, it seems that way in terms of its maturity or its maturity in our space as things go.
Paul Schneider: That's interesting. You talked about personalization. And my thought is sometimes some of these things become popular and then decrease in popularity. And I think of ARVR, became very popular, very hot, and then people realized, yeah, it's really good, but not everywhere and everything. Kind of like mobile, like, hey, just de facto becomes part of the standard, especially when you don't have to think about it, "Hey, do I do this or that?" You know, it's just part of the responsive design or approach. But personalization, I think probably one of the things, because I've seen this happen before, is, yeah, it becomes popular 'cause people realize it's needed but then people realize it's just hard, and maybe I can't prove the value versus the effort. here comes AI and in my opinion, probably making the effort much, much lower. The barrier to entry, if you will, decreased and so they're like, it's being rediscovered, if you will.
Donald H Taylor: And personalization means different things to different people. For some people it's a very intricate analysis of what an individual needs. For other people, it's just, well, if they've got this job title, they need that course. So it's a very loose view here of what personalization means, because it's by no means the same definition across the board.
Chris Van Wingerden: Mm. Yeah. I think of a previous stretch when personalization was very hot buttoned. And of course, the phrase that seemed to float up all the time was the Netflix of learning, you know? recommendation engines. And so anyway.
Paul Schneider: I remember when personalization was, "Wow," it said, "Hi, Paul," when the car started. Fortunately, we have, I think, moved completely past that is, being the barrier there. It was interesting too, you said that Showing value moved up, but that analytics or data was going down. I'm like, most of the time you need to have some data to show that value. So, any suppositions or thoughts as to why maybe that is, or maybe it wasn't a big enough trend and we'll see data come back up later.
Donald H Taylor: No, it was certainly a noticeable trend, and I mention it for precisely that reason. Personalization and learning analytics were in the same spot two years ago. They're now wildly divergent. And the reason for that, I think, is personalization is a clear use, and you can kind of see how that works with AI. Analytics kind of scares people as a word. I think people don't quite know how to do it. They think it's a good idea, but in an abstract way. And the showing value piece, I think that's gone up. When I talk to people about this, they say it's partly, in fact largely probably, because people are scared of their jobs, scared for their jobs, and they think, "I need to show value in order to keep in control of my career." So yeah, it is weird. I think that people are taking the shortcut to keeping their jobs by personalization, personalizing rather than actually using the analytics to link to value.
Chris Van Wingerden: You could see where proving value, one way to do that is reduce costs, which, you know, AI, if you can do more things faster, then you're reducing costs.
Donald H Taylor: I think you're exactly right. I think that's a short-term view of it. I think a lot of people are thinking, "We can cut costs. We can produce more content faster." I don't think that's necessarily the right approach to using AI. I think in a few cases, we need more content. In most cases, we need better targeted content, and that's a different kettle of fish. And, I'm not sure we're seeing people rise to that challenge yet.
Paul Schneider: Yeah. I was thinking with AI, two parallel different conversations. One was, with the AI and it enabling them to do some things that they'd wanted to do, like personalization, maybe better improved, more simulation type things that are not only novel, but people know they work. So I could see how that could be thrown forth as showing value. And then on the flip side, of course, we're seeing a lot of, "Hey, upload the document and the tool will create your course for you." and conversations about here goes a race to the bottom, you know? who can produce the most garbage that the fastest. made me think a little bit of our previous, Donald probably in your surveys, when a PowerPoint converter, was the most popular thing. And hey, wow, now we can... A and of course for a while it was a race to build content faster that was worse, you know, on it, until finally there was some backlash there.
Donald H Taylor: I think that there is, a, not necessarily a race to the bottom because you can produce content that's good enough. And that's the danger of AI. You produce... I mean, certainly there's a lot of rubbish being produced, but you can also produce content that is good enough. And it's not just a bit cheaper, it is incredibly cheaper because you don't need a skilled resource to do it, and you can press buttons and, like you say, produce a course or a job aid or whatever really quickly. And I know Josh Cavalier, other people have shown videos they've produced, which you know if you've ever tried to produce video for learning, how absurdly expensive it is, and you can do it for 1/100 or much less of the cost. Given that, do you really care if the video isn't completely accurate? No. You're gonna spend that money and then make the saving. And I think for when I say it's challenges on the cost front, that's where it's coming from. That anybody can do this stuff now very cheaply, and that puts the instructional designer, I think, in a difficult position. Yeah, I'm thinking when you I'm not trying to bring everybody down, by the way. I'm just trying to say it as I see it.
Paul Schneider: When you talked about creating that video and it made me think of, one, people are like, "Hey, can I just feed this stuff in and it creates the video and I'm done?" Yeah. And then I think of, one of the early courses that, one of my teams worked on, and it was a very engaging scenario with video clips, and it was about insurance, insurance investigation, and it was fun and all different characters and stuff. And it had real actors, coming in and doing that. And yes, it was expensive to do it then, but I was thinking in AI world today, they could certainly have shaved some costs, but if they... I mean, the key to that success was not that, A, the acting was good and the video quality was good. It was, there was really very well thought out instructional design behind that process and that scenario. And, I'm afraid that, as you said, sometimes you go ahead and skip that step too. and maybe AI does some good things on there, but does it really build out that whole piece in that way? Or, does the person building it, have enough experience, in that to go ahead and correct it as opposed to, hey, as you said, good enough.
Donald H Taylor: I think in some cases it will be good enough. The sort of complicated, complex, and emotionally engaging thing you're describing there, Paul, I think is very difficult to do with AI still. The problem is does the person who's buying it know the difference between a good and a bad product? Likely not. And they're confronted with something that costs $100 or something that costs $5, it's a very easy choice for them. And not only do they not know the difference between good and bad, but they don't know the difference in impact because nobody's ever really shown the impact of training successfully at the sort of level of complexity of changing behavior that you're talking about, Paul. And that worries me. That worries me that we're gonna end up with a lot of dross out there, people thinking, "Well, that'll do the job." It doesn't do the job. Nobody's gonna get the blame for having saved money on producing the thing. The person who will get the blame will be the instructional designer or the person who was pushing the buttons of the AI, which is a double whammy, isn't it? You're not getting paid to be an instructional designer and being blamed because you weren't an instructional designer trying to make the thing better. That's tough. We've gotta pull this round. We're delving down into a bad place here. Lift, lift us up.
Chris Van Wingerden: Well, it, but it does highlight something that has been a constant conversation here on IDIODC, is that whole proving value. And we're talking about proving value in terms of end results as opposed to proving value in terms of budget. Uh, you know, the number of conversations we've had over the years about demonstrating that your work is actually having an impact, whether that's bottom line things, reducing risk, preventing injuries, being able to demonstrate the actual, ROIs and the things that we're supposed to be influencing has always been a challenge for us, so,
Paul Schneider: I think, back to proving that value and the analytics, It's not surprising on some level, why the most of the AI focus has been on the, what I call the front end and the creation pieces there. one, that type of AI stuff applies to creation of just about anything. I mean, whether you're marketing, sales, et cetera. So from a business standpoint there, of course they're doing those types of things. And then from a learning standpoint, hey, these things do take time. It's easy and I can, it, this is the quick win in there. And, with analytics it's always been, yeah, I mean, we've been stuck so long, and stuck is a way to say it, but SCORM, with its limited analytics. You know, we got xAPI, and then that opened up the door, but people are like, "Yeah, the door's too big. I don't know what to do with it." You know, it's open to the world. Let's stay inside the house that I know. but that's kind of my, on the positive end, my hope. as you mentioned early on, the AI can really help with the analytics. I... We did another session, and a webinar, about a year and a half ago, and we were talking about xAPI and feeding the analytics in. And I know with dominKnow, one of the few tools out there that actually tracks a lot of xAPI data, we were able to feed that in. And I was able to start to get some answers that before I knew I could get but you'd have to write the reports and nobody did it, and there wasn't any standards of what data to collect that, at least standards that everybody was following.
Donald H Taylor: I love the idea of getting the data through the APIs, and being able to make correlations and show value. Can you say a bit more about that? Because that's very unusual. Typically, people can't get access to the data, or if they do, they don't not quite sure what to do with it. So what did you do?
Paul Schneider: Yeah. So in our particular case, and this was... I just was trying to show that, hey, AI can really help you out, you know with this big data that you may be scared of or don't know what to do with. and so with dominKnow, the defaults and stuff are sending all sorts of things like, what did they click on? Did they went through a scenario? What path did they go through? all the practice exercises, and what learning objectives were these associated with? heaven forbid, we tie the testing questions to learning objectives and practice exercises. But is, uh, and what they clicked on and things. So, or even, like, if you have fill in the blank, you know, what did, did people respond to that? What are people saying? So, all that data was, of course, fed to a learning record store.
Chris Van Wingerden: Mm-hmm.
Paul Schneider: Nothing new there. But then people in the past and there would be, "Wow, I got this data and I don't have a, I don't have a Google Analytics that's forming all these, results for me." And so we were able to download that data, feed it into, in this case, Sarah Mercier and their BuildCapable team had built a little bit of a Chat GPT, widget that understood xAPI better. so it was trained a little bit on that and fed it in. Of course, as you're feeding in and everybody knows about AI, you can't trust every answer there. So I was checking it and yeah, sometimes if I didn't give the right prompt the right way, its conclusions were right mostly, but not right in other cases. So I had to tweak it there. But I was able to get out all sorts of reports that before, unless I had a really good dynamic reporting engine, which you can find in some tools, but not in most xAPI related tools, I was able to get out, a lot of different, results, you know, and see, hey, these people are taking this path down the learning. Uh, this is a word cloud of what people are responding to. Hmm. These are the objectives that, people are having trouble on. And, yeah, it was, I mean, it, it made me hopeful. There's a lot of possibilities here, and it... Not surprisingly, from a product standpoint , we're looking at bringing some of those things in to make it even more accessible for folks, and I think that is, the, the word is, is accessible and either seeing examples and the way it's doing. Like, I mean people that were afraid of coding are doing vibe coding now, so some of that has changed too.
Donald H Taylor: I think that the AI can do a lot to help people with analytics, with data, because you don't need to be a SQL expert. You just need to know to ask the right questions, providing things are set up right. To ask the right questions, and then to ask what I call the second question, the important follow-up question. So your first question is, what happened? And the second question is, why? Why is this number bigger than that one? Why is this trend in that direction? This looks odd. What's happening behind that number? And that sort of analysis and inquiry can lead to very often substantial insights will help you either improve your training program or demonstrate that actually you are having an impact. Yeah, so we've got to a better place now. We've got a, we, we're in a place where there's possibilities happening. That's good.
Chris Van Wingerden: Um, so we've been talking about the hot topics. you mentioned though as well, there was a question about what people see as the challenges. Not that I want to keep us in a negative mood here or anything, but, what, what kind of revelations challenges are opportunities, Chris There we go. Exactly.
Donald H Taylor: There we are. I like that. I like that. The key challenge was... Look, you asked the, you asked the question, what's your challenge for the, the year ahead? It's very interesting. The first so it's been going for five years, 2026, 5, 4, 3, 2. So in 2022 and in 2023, the first two years, 40% of people answered it. And then after ChatGPT came out, so in 2024, so that's the first y first year when we've had ChatGPT, around for a year, 95% of people answered it. And it's been then dropped down to 85, back up to 95. So the first thing to notice about this question is that since we've had public access AI, people have felt they want to share more about their challenges. And the number of words that they're sharing has gone up from 14,000 in the first year to 41,000, which is about the same length as, Heart of Darkness by Joseph Conrad, which of course is the book that Apocalypse Now is based on. Yes. You know, it's, it's fairly, fairly dramatic stuff. The... There are five trends in there. The, the key ones, the... Some are ones you're always gonna find. Budget, budget and resources, engagement and application. How do I get people to come to training? How do I get people to do their training? but there's one which is definitely... Two, two which are definitely new since ChatGPT. One of which is, what's the role of L&D now? And that's a really fundamental one. It's... It appears to be changing to me, but I don't know how it's changing, and that is concerning. So that's a, that's a fundamental challenge people have. The other one, of course, is a more transactional one. I don't know how to use AI, or rather I'm not quite sure about AI and using it in learning. So there's the sort of very specific one about AI, and the very larger one about the role of L&D in this new world. The interesting thing is that when I ask also the question, "What are you doing that's new this year?" The very positive thing is that, and it's... It was a bit odd, the things people were doing that were new seemed to match pretty much what they said their challenges were. And I I couldn't work this out for quite a long time. I was, I was sort of wrestling with this. Why, why are people saying this is a challenge, but also saying they're doing it? And then I worked out, well, it's exactly what's happening. People are recognizing there's a challenge, and they're taking action to meet the challenge. And so if you could... You can look at those two answers for individuals and see exactly that happening. So people would say, "The challenge I've got this year is that I need to really start showing value. What I've done in the past 12 months is I've moved from simply delivering transactional courses to helping people really learn and being focused on what they do at work." Now, that's a big step already, and then the challenge is to take it to the next step, which is to actually show the impact of it. So y you can see that it is... People are making progress and they're being positive about things. If you just look at the challenges, it can be quite depressing. Um, but interestingly, what have I done this year? 38,000 words, almost as many as the challenges as well. So people were as keen to share the progress they've been making as they were to share the challenges they, they faced.
Chris Van Wingerden: And those two questions, those were them typing in information as opposed to selecting from a list?
Donald H Taylor: Free text box. And of course, some people just type in budget. Some people just type in AI. Yeah. but mostly, mostly people are typing a lot. If you imagine you've got... I mean, 95% of people answered it. That's about 3,400 people this total of. Well, they're, they're, they're typing in, typically at least 20 words each, and some people are typing an awful lot more than that. So you get some we're getting some very substantial answers in that.
Paul Schneider: Wow. Yeah, and you had said that that, that increased. It's, almost like, Everything changes ... there's some big changes, and so I'm working extra hard and doing all sorts of perhaps new things that I hadn't done before. Um, as in the past, it was you know, same old, same old for a little while.
Donald H Taylor: Yeah, absolutely. I mean, at the beginning of the survey, what's your challenge, of course, 2022, beginning this question in the survey, a lot of it was coming out of COVID, I need now to shift towards classroom training again. but here's the interesting thing, in those first two years, 2022, 2023, the word human wasn't mentioned at all. As soon as ChatGPT comes out, the word human gets mentioned more and more and more, and this year it was mentioned 80 times in the answers we got. So people are increasingly wanting to keep the humanity in what we're doing. There may also be a sense that I'm aligning my identity, my human identity, with what's going on. So there's that side of it. The other thing of course is that... And I do think that people use this question as a way to just say, "Thank you for asking, but nobody's asking at the moment, and I'm gonna tell you." Um, the word pressure was reasonably high as we came out of COVID. It dropped down, and then it came up again, and at, it's at the highest it's ever been. And I choose that word rather than stress or overwhelm, which have shown the same pattern, because the word is always used by It doesn't matter what language they're speaking. You translate it into English, it comes through as pressure. But it's always the L&D departments under pressure, and that's increased each year. So I think that combined with that sense that the word human has come up, combined with a sense that people are just sharing more and more of their problems, I think I'm getting this real sense that people are concerned. But as I say, when I look at what people th say they're doing for new, eh, I'm optimistic about the future.
Chris Van Wingerden: Mm-hmm. I was thinking, and you just, you mentioned it, but it, it almost felt with that increase in volume of response, et cetera, that it's like almost like someone... The... So many folks were just saying, "Oh, thank God, someone's finally listening to me," or, or, "I'm finally having a chance to express these things that have been rattling around"...
Donald H Taylor: I think sadly, if you look at, you know, most people work in very small teams. So it's one, two, or three people in L&D, and so we are all i, in our own little bubble. And one thing I absolutely say to people is whatever you're doing, just network and talk to people about how you're feeling about things, because there's a real sense that people could get burnt out. And literally just before we came on air, I was looking on my phone at somebody who's talking about she's a sh she's a, uh, a freelance trainer, a freelance instructional designer in the UK. She's, uh, got two teenage sons, she's got a mortgage for her house, and she's saying, "Look, this is my shopping this week. Got any tips on how I can save money?" Now that's really fundamental stuff. Mm-hmm. And, I don't want people to be in isolation where they're going through that. We should be trying to do what she's doing. Let's share with other people and let's get support for each other so that we can help ourselves through what I think is a temporary period, but I think it is going to be a tough period.
Paul Schneider: Yeah. I was thinking it, it gets back to some of that, I need to show value. And I, I think it's I don't remember a time in my career where showing value wasn't important, but there, there's definitely been periods where it's like, "Oh, if I don't show value, we're on the chopping block," type thing. or even the existential of, "Where do I fit in? This is my profession, where does it fit in there?" which I think also shows some of the human, aspect of it and bringing that in. Yeah. I've seen the human aspect brought in and commented in all sorts of other industries when they're talking about AI. And, sometimes it's brought into, the whole phrasing of AI is not replacing you, it's working with you. But, obviously people have seen pressures, and you see the people saying, "Oh, we're laying off these folks because of AI," which if you kind of look at the numbers, it's because the business isn't doing well. They're just trying to come up with a positive, what they see as a positive spin. So, yeah, but you know, people read the headlines, they don't read necessarily all the details underneath that. I think on one hand the pressure's not a bad thing. Pressure is what produces steel, produces all these... And, and causes us to change. And actually, you know, back to maybe analytics, hey, this is a great way to fight against slop or, or things that really aren't well thought through that don't have someone with that experience guiding it.
Chris Van Wingerden: Mm-hmm.
Donald H Taylor: Yeah. I think that's a good way of... A, a good reframe.
Paul Schneider: I'm curious as what are some other things that were really hot and then just completely faded away? 'Cause sometimes we get caught up on the what's here now and get all worried, and then it's, it's a non-issue in a year, so.
Donald H Taylor: the classic one of these is microlearning. And I'm not saying microlearning doesn't work, I think in the right conditions it can do. Depends how you define it. But in terms of it being something which received a really noticeable injection of cash and then hype, microlearning stands out as something which, in 2015, it stood in a particular level on the survey. 2016, a bunch of companies in the States got funding. 2017, the vote in the States was very high for microlearning. Just it was, it wasn't high at all in the rest of the world. 2018, it started dropping in the States, but it took off in the rest of the world because the United States sneezes, the rest of the world catches a cold. Things ripple out. And then 2019, everything was on the slide again. So that was a classic case where you could see that a particular marketing effort had been made around one particular category, it had exploded, and it started in one place. We felt the ripple across the world, and then it went away. Uh, I'm not saying it's wrong. All I'm saying is that with that you could see very clearly what had happened. Now, the same thing would've happened with the metaverse. I put the metaverse on, I think in 2025, and I was a year too late. If I'd put it on a year earlier, it would've absolutely spiked. but by the time I'd put it on, everyone's attention had shifted. I put it on in 2024. By the time I'd put it on, everyone's attention had shifted towards AI and the metaverse was nowhere, and it's now firmly at the bottom of the table. Nobody's voting for it. Uh, but we've seen these things spike and fade away. That hasn't happened with AI. AI was first on the table in 2017 as a choice. It went up, it came down exactly as I was expecting, and then Chat GPT comes out, the whole thing goes through the roof. I'm actually taking it off the survey next year because firstly, it's very boring if the same thing is number one every year. Mm. But secondly, this is not the same as everything else. You can't say, "Do you prefer microlearning or AI. It's a bit like going to your kitchen and saying, "What's most important: microwave, dishwasher, hob, or electricity?" Doesn't make any sense, right? AI is the electricity that powers what we're doing, and, we just have to accept it now. And actually, the, the latest report from me and Edgar Vinasco, which is this one, AI and L&D, that's from September. That's, that was just before the release of the, of the Global Sentiment Survey. That showed that over half of the people we'd surveyed for that, which is over 600 people from 50 countries, were using AI regularly in their work, which was a big jump from the previous year of 40%. So it's gone from 40 to something like 53%. It's just part of what we do now. It's part of our lives. And so I'm not gonna... I'm gonna take it off the list, and I'm gonna spend a lot more time looking at the words that people give me to describe where they're going because there's so much richness there, and I really feel I need to honor the effort people put into sharing their thoughts and their feelings about the challenges and the successes.
Chris Van Wingerden: Mm-hmm. for that richness, how have you, you know, to this point, dug into that? Are you reading it all and, and creating your own word clouds, or are you using AI as part of that? That'd be interesting to know.
Donald H Taylor: Chris, great question. And look, for a long time, I was a purist. I said, "No, I'm gonna read it all. I'm gonna put everything in, translate it. You put, frankly, in Excel spreadsheets, Excel spreadsheets, and you have this list of categories, and you're ticking them off one by one." You can do that up to a... After a certain quantity of stuff, it just becomes impossible to do it all. So I still do the reading, I still do ticks in the boxes, but that's not the only thing. I then go to, actually, two different engines, Chat GPT and Claude. And I say, "Right. Here's the data. Go through. I'm not gonna tell you what I think my categories are. You tell me what the categories are," and then I do a comparison, I see between those two and the data that I've got. What have we got? I interestingly, it's a pretty good match most of the time, and I then have to... A sort of subsection that I have to go and make a choice on myself. and it usually throws up some anomalies that I've missed or some new ideas. So that's very helpful. It's a combination of the old thing, Chris, a bit of, a bit of technology and a lot of sweat still, I'm afraid. At some point in the future, I'll have a robo Jeeves who will do all this for me, and I don't have to, I don't have to worry myself.
Chris Van Wingerden: Well, you could probably pick up a, a, a website URL Ask Jeeves, 'cause I think that one folded, in the last couple of weeks didn't it?
Donald H Taylor: Is it last two weeks, was it?
Chris Van Wingerden: Yeah, they finally
Donald H Taylor: Do you remember... Yeah, they finally threw in the towel.
Chris Van Wingerden: Yeah, they did. They admitted defeat to Google finally, I guess.
Donald H Taylor: Didn't Jeeves... Wasn't there a Jeeves character leading the New York New Year's parade in New York one year.
Paul Schneider: bet you there was ... how much It was,
Donald H Taylor: there was ... It was,
Paul Schneider: it was... They had a lot of money behind it for for a few years. Not like any of the, you know, "Hey, let's deliver your pet food for free online," Yes think not thinking about the weight, you know, and things like that. They had a Super Bowl ad, and a year later they were out of business, you know?
Donald H Taylor: I'm gonna look that up right now. I'm sorry, I can't resist it. Jeeves and the New York parade.
Paul Schneider: this brings it back to me, the, some of the different things. You talked about mobile, a and AI has become like mobile. It's just part of the ecosystem now. people are expecting to go ahead and use it. They're already using it at least on some level for different things. in fact, I, I remember, a friend who, they interviewed someone and they said that they didn't use it at all, because they were a bit of a purist. A and that was actually a turnoff, for them. They did not hire that person, even though he was very well qualified in many other areas, 'cause they're like, "You know, you can't just stick your head in the sand," so to speak, was their perspective on it.
Donald H Taylor: I misremembered. It was the Macy's Thanksgiving Day Parade between 1999 and 2004, serving as a prominent internet era icon. Ironically, that s answer was provided for me by the AI overview in Google. So I think that's the sort of meta message there for poor Jeeves.
Paul Schneider: Yes. Was, be, be ahead of its, ahead of its time perhaps. You know, that's, it's a good point. The technology has to land at the time when people are ready to use it and to trust it, and AI is kind of there, kind of not there. I remember in 1999 having a conversation with somebody from Microsoft and him saying, "Yeah, this is gonna be the year of e Christmas." I said, "What?" "People are gonna buy their stuff online." Now bear in mind that Tim Berners-Lee had put forward his proposal for the World Wide Web 10 years earlier, and we'd had a whole decade of the '90s trying to get the World Wide Web off the ground, and it'd been creeping along. It'd been used for email and what have you. It kind of made sense from a technology point of view that it would be e Christmas, but from a user adoption point of view, we weren't anywhere close to it, and it was a fraction of a percentage of purchases online in, in Christmas 1999. Now of course, i it's an absolute de facto. It just takes time. What I would challenge everybody listening is to think, well Where are we now in that same story of the World Wide Web, if you can think back that far?
Chris Van Wingerden: Mm-hmm.
Donald H Taylor: Where are we now? I think we're right at the beginning. I think we haven't begun to see all the impact we're gonna see with AI on instructional design, on learning, on everything we do in our lives. And, and there are things which are going to happen which are unimaginable, and which will the stuff we're doing now, in the future, will look as quaint as Jeeves being the Macy's Thanksgiving Day Parade looks to us now.
Paul Schneider: Yeah, you mentioned about the internet and getting to that shopping point, and I, I was thinking, you don't think about it, but one of the biggest drivers of making that to that first step was the success of AOL and making it easy for people to actually get online. because internet was there, but y if you weren't at the university or, or something, y just getting to the highway, so to speak, as they, framed it back then, was hard. And we talk about courses people are doing, "Hey, here's the certification in AI this or certification in that," and, and, getting, trying to help people get to that next stage. Same time, we're also seeing, I think even more rapidly of tools having things built in that you don't really need to learn anything, you're using it to help you do what you could already do, but it does faster. Mm-hmm. I think sometimes we run in the risk, and this is where that training comes in, of you're using it to do something that you have no idea how to do. And so when it does it, it, it looks, it always looks pretty good, but it isn't always pretty good.
Donald H Taylor: But Paul, I think this is a really interesting point. Do you think there's a risk there that we let our skills atrophy, and that we start effectively flying a plane with automatic controls and we don't quite know what's happening with all of the stuff on the plane? Uh, we kind of hope we don't crash it, but everything in the cockpit looks great. Are there risks involved here?
Paul Schneider: Yeah. that's, definitely a risk that I, I think about. Flying a plane is a great one because there's so much training someone goes into. Yeah. At the same time, we know those pilots are using a lot of auto technology, but we trust that if something goes wrong which is almost never, but people start, "Oh, it's never." No, it's not never. That they can take over and do things. And we've seen, fortunately, time and time again, where they have taken over. And because of their training, the training of the a airline, attendants in helping out and stuff, things have happened. You know, we have all these things in place where we, do we go ahead and say, "Hey, we don't need someone who really knows this sum. We just need somebody to push the buttons." in, in the pilot area we've said no. That that's too high risk. Yeah. But I mean, yeah, it could be a risk in training, and that gets back down to showing that value, and I think, hate to bring it back around to, but that analytics of, you know, what is actually the training that's making a difference.
Donald H Taylor: And I think this means that we've got a responsibility as people who are involved in L&D, whether it's as instructional designs or anything, to really understand the business and the impact of what we do and what people are doing in their jobs. And if we understand all that, we're in a much better place to be instructional designers to do analytics. I think we have to just get a bit closer to understanding the jobs that people are doing to enable us to do that, whatever else it is we, our role is.
Chris Van Wingerden: And a related but slightly different risk too, thinking about some of the things I'm reading in, in the programming world. If all of the junior programming roles are being replaced by vibe coding and AI, then how are we ever going to have senior, programming minds to be able to actually understand what's going on and what should be done? So there's that, that continuity of, of the, of the craft of the knowledge base too that, that could potentially be at risk too.
Donald H Taylor: I've been looking a lot recently into how people are learning in the legal profession, and of course the legal profession is going to be hugely impacted by AI, and it already is, because it's all about words. And they have these incredible applications now which y you can get the application to draft contracts based on the general knowledge of case law and that company's particular IP in a way that, in the past, a bunch of juniors would have gone off and done the discovery, gone and read stuff, produced documentation that would enable you to say, "Well, this is the right contract for you, Mr. Client, Ms. Client." Now, if they're not doing that, these junior partners, what are they doing to learn the trade? I was talking to Heather Stefanski at McKinsey, which of course is a completely knowledge-based industry, and she's very focused on... She's the chief learning officer. She's very focused on helping people through their career learn how to do their jobs. And so what they're doing is they are building learning into the tools that people use. So you don't just press a button and something happens, but you work with a tool to make something happen, and through it you are building your understanding of the process Without necessarily having to, in the case of lawyers, read every document. So it's like an analogy. You're doing the work, but perhaps more at a distance, but you're still learning the process without having to go through the pain of having to actually read all the documents. That's one example, and I think we have to be much more mindful in the future, to your point, Chris, when we've got new people in knowledge industries, of saying to them, or building systems which enable them to learn in a way that is intentional rather than we just say, "Go and do this job. Figure it out, and by the way, on the way you're gonna learn some stuff." We have to be much more intentional now about structuring their work to make sure that learning is a part of it, and that they don't do stuff without understanding the process.
Paul Schneider: I was actually having a conversation yesterday about lawyers and how that was changing some of those things, and, and they were getting bit because they didn't actually check their work and do things, and sometimes it wasn't right. But, I, I think we forget sometimes learning's hard . No matter what it, it... No matter how good it is, it's still hard to, to learn. I mean, otherwise people would be going to school all the time. There wouldn't be spending that effort. And I love how you've had, that they were reframing it of, hey, a and this is hard for companies, go, "I don't just need it done by someone to hit the button. I need it done and I need them to start to really understand what done means, how it got there." And so, you know, when different things or changes, they can adapt to it. They can do a quality check, so to speak, on there. And I like the idea of, again, the company has to do it, and that gets back to showing value of that, "Hey, I can hit the buttons, but I also know what's happening when I hit the buttons and why. And yeah, I can't calculate it as fast, or I can't write it as fast, but I can tell when it's wrong or something is funky looking or, or something wrong there." And kind of building that in. We always talk learning in the state of work. It reminds me of kind of way back when, before there's any of this, y you did a journeyman. You know, you were learning on the job and, and just, learning from that expert, there and actually going ahead and doing it. but then someone is helping you and, and forcing you to go ahead and question, not question badly, but question goodly, about what you were doing and why that made sense and why you were doing it. And you weren't just doing it 'cause someone told you you learned a procedure.
Donald H Taylor: Yeah, exactly. I think we're gonna be going back to that sort of apprenticeship journeyman a a approach much more. And again, actually interestingly, it's, it's a word that McKinsey have used for a long time, apprenticeship, the idea that you consciously are learning your craft as you go through.
Chris Van Wingerden: Mm-hmm. Well, we have covered a whole range of things. Donald, so glad that you were joining us here today. and we're looking forward to the 2027 version of this survey to catch up, on, on what's gonna happen in, in the timeframe. Every year just seems to go a little quicker, doesn't it? It does. It
Paul Schneider: does. I'm surprised you didn't say it's a journey that we've been on.
Chris Van Wingerden: Ah. Oh my gosh, I missed a bad pun?
Paul Schneider: Uh.
Chris Van Wingerden: I know. Ugh. Anyway. Um, hey folks, instructional designers in offices drinking coffee, drinking tea, drinking hot chocolate, drink not drinking at all, whatever, it is brought to you as always by the team here at dominKnow, makers of dominKnow | ONE, helping L&D teams to develop, scale, and deliver learning that maximizes employee value. Check us out at dominknow.com. Donald, thank you so much for joining us. As always, a great conversation, really fascinating. and, and we've... As I said, we look forward to hearing about next year's version too.
Donald H Taylor: Can't wait. Always a pleasure. Thanks.
.svg)
.svg){kind=icon}
