Are Your Employees Ready for AI? | With Katie Jacobs

Speaker A: So my guest today is Katie Jacobs. She's a journalist who spent around 15 years writing about work. Um, she's worked at the CIPD and now writes for a range of HR and business titles. She also writes and contributes to long form research and reports for a number of other organizations and institutions. And over the last year or so Katie has spent a lot of time writing about the impact of AI in organizations. So that is what we'll be talking about today. So if you've bought a new tool and you're wondering why half your team seem to be working around it rather than use it, Katie is going to help us find the answer. Uh, uh, thanks for being on the show Katie.

Speaker B: Thank you very much for inviting me. I've spent the most of my time, I feel over the last year kind of researching, talking about, writing about the impact of AI on work and HR in um, in particular. So excited to have a chat with you about some of the things that I've learned. Although obviously I have to caveat always. I'm not a technical, I'm not a technical expert, but the kind of line between where AI meets work meets people and culture I think is a really, really interesting and rich topic right now. So thanks for having me Sarah, for sure.

Speaker A: So as we said, you've been writing about um, workplace technology and the impact in organizations for around 15 years. Um, what do you see that's different now with the um, rise of AI compared to other waves of technology in the past?

Speaker B: I think the thing about um, AI and I think when we say AI to kind of uh, classify what we mean, I'm talking about generative AI, is that it is, the speed at which it's been adopted is just so much faster than kind of any other technology that I've seen in the past. And I think that is because of the kind of, the accessibility and ease of use. Um, I was reading something over this weekend, um, that reminded me that ChatGPT is the fastest ever adopted um, kind of technology in consumer history. Um, it quickly became, I think it had 100 million users within like a month. And it was just released almost as like an experiment to see what was going to happen. And within like a month like millions and millions of people were using it. So there's something about that kind of accessibility, how easy it is to use. You don't really need to understand the, how the tech works. And I think the, compared to other workplace technologies, definitely the kind of line into our personal lives and see how people, why people will be using it to make their personal lives easier. Whereas I think when stuff like potentially like email and stuff came along, maybe people weren't so keen on plugging that into their, into their personal lives. Obviously is part of all our personal lives now. But I think given we're also used to using technology, being able to, you know, ask it literally any question in the world and get an answer. That's speed. And I think we'll obviously come onto this. But what I've seen from my conversations was like chief people, officers and um, HR leaders. Basically people were using it before organizations had even kind of figured out they needed a policy or some guardrails or governance around it. It was kind of the horses already bolted. Um, and like you're seeing this with the regulatory stuff that's happening now and all the stories we were talking about before we pressed record about how it's going to kill us all and regulators can't keep. Yeah, the regulation can't keep up because the technology is so fast and the organizations can't keep up because the usage, like the adoption was so quick. And I think that is, that is what makes it feel markedly different. And that's why. And you often hear, don't you, that it's like going to be, have a bigger impact as the Industrial Revolution, but on like a faster, at faster pace. Uh, and I think that's probably why it feels so different, but because people can use it so easily. When you start using it, it's very clear how it could do your job potentially better than you or at least

Speaker A: faster than you, or where at times when maybe you don't want to be doing your job, like. Yeah, it doesn't get tired.

Speaker B: Yeah, exactly. So you can see that, uh, in like a positive way or you could see that in a terrifying, like anxiety inducing. Oh my God, sometimes both. Yes. Am I going to be like, obsolete? Because like, particularly the kind of models that are coming out now, you're like, oh my God, I can't believe that it could do that. And it did it in like half a second.

Speaker A: Yeah, yeah, we actually, we've been trying to build some dashboards here in my business and we've been going back and forth, back and forth on what tool we're going to use, et cetera. One of the team just put it all in code, all the data in Copilot yesterday, and it produced exactly what we were looking for. And I don't know, we use Copilot all the time, but we just hadn't thought about it and now we've seen it we're like, like the capability. The possibilities are uh, endless really. But, um.

Speaker B: Yeah, yeah. And the possibilities are endless, I think, in both directions.

Speaker A: Yeah, yeah, yeah, yeah, exactly. Yeah. That's what I kind of stopped short because I'm sure we'll get into kind of the. Some of the pros and cons. So when you're talking to businesses, is your impression that in the main all businesses are considering um, AI adoption as part of their business planning? Or are there some businesses that just really, uh, keeping their eyes shut and not really interested, hoping it goes away?

Speaker B: I think this comes back to the speed point because I think if you'd asked me this a year ago, I would have said I don't think all businesses are. And kind of an anecdote I always use is kind of about. I think it was about this time last year or towards um, Christmas last year I was working, uh, um, with a major consultancy and we were doing lots of work, kind of thought leadership around how AI is reshaping work. And so I was very much down that rabbit hole of all these amazing stuff it's going to do and it's going to total you to rebuild work from the ground up and redesign organizations. I was doing that and then the next week I went to a conference. I was working at a conference of public sector HR people and one of the speakers asked how many people are using ar? And not even half the hands in the room went up. Um, and just a kind of enormous scale, the difference, the variance between kind of high adoption drunk the Kool Aid, have loads of money to invest and low adoption have the money. Because I think one of the issues here is people say it's going to make everything. Oh, it can make you super efficient and you could save, you know, lots of money on headcount. But actually the investment you have to put in the first place is it's not small. So actually, you know, people are finding they can't afford to build the foundations in order to use it. Um, so I think the scale of adoption, but I think now it would be. I'd be very, very surprised if there are any business leaders that are not thinking about it in some way. I think the thing when we talk about um, Gen AI or how I think about it, it's almost on like two levels. You've got this personal productivity usage and I guess that's how individuals are using it at work to help them work faster will come on to, I don't think necessarily work smarter, but maybe some people, um, and then you've got the kind of, the bigger picture, reinventing how work gets done. Totally re engineering workflows, almost collapsing bits of organization into each other. And that is I think a bigger picture, uh, wholesale work and org design piece that not all organizations have the appetite for or the time for. And obviously there's like a whole amount of unpicking, um, you need to do to get into that. Um, so, uh, almost you've got kind of operation on these levels. And I think a lot of the stuff we see around McKinsey stat, that always gets rolled out about how 95% of AI pilots aren't delivering any ROI. Yeah. She says I have not looked that up. So I might be wrong, but I'm

Speaker A: pretty sure it's their word for it for the moment.

Speaker B: Used it a lot. Um, is because it's kind of stuck in this kind of disparate pilot mode where the pilots aren't collecting. So you're not able to generate value because. And I think that's because the um, actual work that needs to get done is so massive that people almost cannot think big enough or are too scared to think big enough. Um, and maybe it's not the right thing to do to think that big as well.

Speaker A: Yeah. But on the um, percentage of failure point, because it is a new technology, there is an element of just testing and letting stuff fail, seeing if it works. So I mean that does sound high, uh, and like a big investment for businesses. But I suppose we shouldn't be completely surprised that there's a higher proportion of failed AI projects purely for that fact. Because it's new. We're all figuring out how we're going to use it and what we're going to use it for.

Speaker B: Yeah, I think so. I think maybe it's not that they've failed, it's just that they fail to deliver value.

Speaker A: Right.

Speaker B: So maybe they're. So I think the question isn't necessarily are they working, it's are they actually bringing value to your business?

Speaker A: Is it worth it?

Speaker B: Is it worth it? Is it worth like the time and energy you've poured into it? Uh, is it worth the potential negative outcomes or unintended consequences that I'm sure we'll come on to?

Speaker A: Or even the cost, even the cost, um, of. Yeah. Just running these things. And we often we look at, uh, using AI for different. Managing different processes. We kind of weigh up, is it cheaper to use a person to do this or is it cheaper to automate it in some way?

Speaker B: Yeah. And I think you're seeing organizations now I've seen quite several headlines about organizations, um, that you know, got rid of people and then they, and then they've maxed out all their tokens. And I kind of liken it to putting in. If we're talking about other tech again, you know, when you like putting an hris and you've like invested in, to know workday, other systems are available and you've um, invested and then they're just going to keep layering on. Oh, you want to access this, you want to be able to do this, you want this functionality. Well that's going to cost you. So you must, you just become in the pocket of a different tech provider and that might not necessarily bring you the productivity or efficiency or cost savings that you think it's going to because ultimately you don't own anything. Um, and I think that is an issue as well.

Speaker A: Yeah, yeah. You mentioned earlier on about um, you were somewhere and you asked people how much they're using AI. And I had a similar experience at an HR event. Somebody asked the same question and maybe like 10 people put their hands up out of this whole room to say they're using AI. And I do sometimes wonder about the governance side of things and as particularly in a larger business, in smaller businesses it's kind of different. Um, but if you've got, if you're in a bigger business and there's lots of governance and you've got a big HR team with kind of lots of, of with like a big risk management approach, then it must be quite difficult to kind of get stuck in and just be testing and trying different things. Have you um, seen any, anything, any examples of that in your work?

Speaker B: I think on the governance side it's what the point I came back to at the beginning is that adoption got advanced so quickly that organizations were always going to be kind of playing catch up. And you know, we all, I think it's called like shadow AI usage, um, which sounds kind of sexier and darker than it really is. I think it's literally just using AI tools outside the realms of. So I don't know if you use Copilot as an organization and they're like, well I prefer Claude, so I'm just going to use Claude. And they're putting your company information into an external model. And I think that is happening absolutely all over the place.

Speaker A: Definitely is because I've heard anecdotally of grieve, um, disciplinaries and things happening for that. Exactly. That, um, putting confidential sensitive data uh, into those tools.

Speaker B: Yeah. And I think that is, again, it's the, the adoption running away before the governance kind of gets its shoes on. But also it's the fact that, um, this always happens, doesn't it? Like, organizational tools often aren't as effective as the tools that we can access as consumers, or we don't enjoy using them as much, that the user experience perhaps isn't as seamless, um, as we want it to be. So I, Yeah, like I said, I think a lot of organizations have kind of been playing catch up with putting the rules in place and the guardrails, um, because you don't. It almost feels like social media. Like, do you remember when I got, like, my first ever office job? Well, my first, like, proper office job, when I think my first journalism job, I wasn't allowed to go on Facebook at work. Yes, yeah, yeah. Um, yeah, it was like a huge thing. And now obviously, like LinkedIn or Instagram or TikTok, you know, those would be part of people's work. So it's almost. It reminds me of how we were very quick to, like, clamp down on social media and then had to row all of that back. And it feels a little bit that. Like that. But I think, like I said at the beginning, the horse is kind of already bolted. Um, you need to be putting policies in place and they need to be, you know, people need to understand what the kind of guardrails are. But ultimately, you can't force people to adhere to a policy either, can you? You probably will be having to, you

Speaker A: know, take people to manage the implications.

Speaker B: Yeah, manage implications, kind of manage performance. And that would potentially be doing more disciplinaries. Interestingly, because I write a lot about, um. I'm sure you've talked about this all the time, about the number of, like, AI grievance, all of those kind of things. And I was writing about that and one of the people I interviewed was like, well, actually, you need to tell your managers that if they take an AI generated grievance and then they put it into an AI, then you might need to discipline them because they've been putting sensitive information into another one. So it's almost like compounding the chaos and creating this, like, terrible death spiral downwards. Because, um, we're all using it.

Speaker A: Yeah, I think it's complicated, isn't it? There's a lot of pieces to consider. And I think. I mean, correct me if I'm wrong, I imagine that the organizations that are more open to taking risk generally in the way they do business are the ones that are going to Be testing more and maybe progressing their uses of this more effectively. But the downside is they probably are also making more mistakes. Um, and that also needs to be managed.

Speaker B: Yeah, yeah, I think so. And it's also, yeah, the kind of capacity for any kind of innovation. Right. And any kind of experiment. And I think a lot of it is down to the, uh, I think the problem with the, uh, with a lot of this is go back to those kind of two levels, like the kind of individual level and the organizational level is how you translate this kind of individual appetite, whether people want to use it or not. And if they're, you know, some people are loving it and they're going off and creating, Writing their own agents and creating kind of their own little bots that are going off and doing their job for them. And how you then scale that and embed that into organizational workflows, that's a challenge. Um, and I just think there's just all sorts of unintended consequences that are just coming kind of being on earth. It's like every time you do something, you pick up a rock and you're like, oh, my God, I can't believe. Like, I was reading a, um, Sarah o', Connor, who does brilliant work on this, um, writes for the ft and she. I was reading a column, I think it was from last week that was saying that one of the consequences that she's observing is that people are taking on work that is outside their kind of area of professional expertise.

Speaker A: Oh, that's interesting. Yeah.

Speaker B: Uh, they're like, well, I can do this myself because the AI can kind of do it for me. And then that is almost destroy. Not destroying is maybe a bit much. It's negatively impacting collaboration because people are like, well, I'm just doing this myself, so I don't actually need to go and talk to somebody who works in marketing because I can just write my own marketing copy. I don't need to go. I don't need to go and ask, ah, for help from it because I can write my own code. And, um, so kind of like weakening the kind of collaborative ties within an organization. And then also people are just getting annoyed. It's going to annoy you. Right. Like, you work in HR when somebody's like, oh, I could do that. I'm a writer. So, yeah, just write my own stuff now. Yeah, it's. It kind of weakens the collaboration and then also kind of has this negative impact on interpersonal relationships as well, because people almost just get into their own little. I can do it Myself, I don't need to talk to anybody else. I've got my army of bots that can get up and do it. And I think those kind of impacts it could have on culture more broadly are probably, uh, overlooked because nobody's had to think about them yet.

Speaker A: That is a really interesting point because I see that behavior a lot in small business. Businesses, obviously, they often don't have big, functional, big functions. They don't have an accounting team, an IT team. So they will do exactly what you're described, you've described. But it, it feels like a good thing because you would. Otherwise that wouldn't be done and they wouldn't be able to afford to do it. So at least it's getting done. But, yeah, I hadn't considered it from that side, where actually, if, if you're not collaborating with people, if you do have those functions and people start to do all the work themselves. And also I would say the quality, uh, in smaller businesses, you sort of accept it as well. It wouldn't get done otherwise. But sometimes, you know, they may have used an AI to write an employment contract. And we'll get it and we'll look at it and we'll think, oh, no, that's. That's not great. It's wrong. It needs to be corrected. So if that's also happening in businesses where they could have gone to the experts, that is creating its own risk as well, isn't it?

Speaker B: Yeah, absolutely. And you're getting like, um. Uh, I think there's two further angles. I wrote a piece that's in the latest edition of, um, work just CIPD's journal for fellows, that's looking at kind of all what it's doing to our brains. And one of the things I uncovered, which I found really interesting, was its impact on how we feel about meaning. And basically, if a tool can do your job, I'm not going to say better than you, but maybe like 80% as good as you. Yeah. How does that make you feel about your own worth and the kind of. The meaning that you derive from work? So I think that's a kind of a big piece. We all just kind of want to be sitting there and you talk about. People talk about, like, human in the loop. But ultimately, if we're all just going like, okay, yes, yes, yes, I agree, like, not. Not great.

Speaker A: How much are we contributing?

Speaker B: Yeah, yeah. And I think so there's a kind of underexplored impact, I think we're going to find about how meaningful people find their work. And I'm seeing, you know, just among my friends, even people who have jobs that are now don't do involve using a lot of AI are actually retraining to do things like with their hands. Like one of my friends is works in animation and has done for about 20 years. And so much of her work was becoming automated that she's retrained as a florist. So she's kind of doing both on side, which I need to, I want to be able to use, kind of use my hands and use my creativity in a, in a different way. So I think you're going to find people looking for meaning in a slightly different way. And then the other angle to it is obviously this kind of AI slop phenomenon that, you know, AI is generating like volumes and volumes of work and it's creating these. Like one thing it's not great at is brevity. Right?

Speaker A: Yeah.

Speaker B: So it's like creating these like pointlessly long reports, pointlessly long things. And somebody was like, well that looks fine, I'll just send it to somebody else. And all you're doing is moving the work.

Speaker A: Yes.

Speaker B: Downstream or upstream or somewhere else in the organization. So it's kind of making your life easier, but you're making your colleagues lives harder because they now have to read your slop and.

Speaker A: Correct.

Speaker B: Turn it into something actually useful.

Speaker A: Yeah, we spoke. Yeah. A few months ago when we spoke, we were talking about this and actually I really have kept that in mind when I use AI to do certain things because actually it's not only about giving it to your colleague, it's about. Yeah, I can produce a draft really quickly now, but the refining of that draft, if I timed it, my guess would be it's taking me way longer. But I think overall there's a good chance that, that the work is taking me the same amount of time. It's just the point at which I'm spending that time has shifted.

Speaker B: Yeah. And it's like, are you spending that time on the value added stuff as well? I think is interesting? I mean like, so in addition to my journalism work, I do a lot of like corporate or brand work. I do research, I do a bit ghostwriting and I'm getting so much more kind of the job is almost like, here is some AI copy. Please make it sound like it was written by a human. It is so much harder to make something sound like it was written by a human than it is to make something that is a bit messy, sharper and tighter and just better. So you're. Yeah, I don't I think it's almost like a false economy in quite a lot of ways. Um, and I think we will see people kind of pushing back against that or almost. And it comes back to this collaboration piece again because will people, you know, want to work with you if they're like, well, all this person does is send me like a 20 page report that they haven't even bothered to read because half of it is this rubbish and I don't want. And now I have to wade through it and that's going to take me twice as long. Or maybe I'll just feed it into Claude and be like, give me the top 10 points.

Speaker A: Yeah, yeah, that's exactly what would happen. If you're enjoying this episode, make sure you're following the show because we've got lots of other episodes coming up with guests talking about different areas of AI adoption in organizations. So going back to use of AI in organizations, um, earlier on this year there was a survey, it was the Writer and Workplace Intelligence Survey where they talked about the fact that nearly a third of workers admit to actually sabotaging the company's AI strategy. Um, what's your observation of that?

Speaker B: I think that sounds about right. Might be higher. I think there's a really high, I think an anti AI sentiment and I think that is becoming higher. I think it's becoming louder. Uh, I think it's, um, I think some organizational leaders perhaps are uh, out of step with how kind of average person might be feeling about it. And I think especially younger people, if you're reading every day like, you know, graduate jobs are, uh, you know, down by half or whatever, they're not. Nobody's employing young people. The junior roles are what's going to go. And so I'm sure you read, I'm sure everybody listening to this is, you know, all those stories from the US about people, you know, execs at Google or Microsoft going and giving commencement speeches for graduates and they're getting booed the minute they talk about AI. That's a massive sentiment thing. I was running a, um, roundtable for chief people officers, um, a couple of months ago and we were talking about AI's impact on early career pathways. And people there were saying they were noticing from kind of their younger employees a bit of an anti AI sentiment. I mean even like I was in the queue for an airplane the other week and there were some boys in front of me, they were probably like early 20s. And they were having a chat about this and one of them said to the others, he was like, well, they're asking me to spend lots of time on using AI and to like train it. But why would I train something to take my job? So I'm just not going to do it. Yeah, well, yeah, yeah, I can see how people feel about that. I'm just gonna, I was just reminding myself of some data. There's a report out from um, I think it's a responsible AI institute in King's, um, King's College London. Um, that shows that the UK public feels increasingly cautious, fearful, negative about AI, particularly regarding jobs and privacy. So about half, just under half of people feel kind of scared about it and prefer to avoid it. And only a quarter view it as positive for humanity and that people are worried about the impact on jobs and stuff. Um, and I think unless organizations kind of take that seriously, then you're just going to have a total kind of chasm between how your people feel about it and what you're pushing for as a leader. And because so much of this has to be grounded in kind of trust and people and culture, it's almost like AI enablement at scale is a, it's almost not a tech project, it's a cultural transformation.

Speaker A: Well, that's exactly where my head was going because um, I was actually thinking about. It's not really related but in my head it was. I don't know if you saw, um, but with everything that's been in the press around the end of humanity because of AI, on one of the um, news channels they, somebody created some sort of AI, an AI image of a person that they interviewed and they asked it questions and, and they started off by asking like a scripted question that it was expecting and they um, then went on to non scripted questions and this kind of AI was responding and they every. The commentary I saw and it was, oh my goodness, this is like, this is amazing. But I just looked at it and thought it feels inauthentic. I know that's an AI generated and, and you could see like everything about its movements. The way it spoke, the way it blinked, it felt. Um, and if that's in an organization, I know that's kind of an extreme example, but it sort of plays into if in your organization you're really relying on AI, you do need to think about how that makes your organization feel to the people that are working in there.

Speaker B: Yeah, I mean there's a lot of um, some stories weren't there about like Mark Zuckerberg wants to create a kind of uh, almost an avatar of himself. And um, I can't remember it was a executive at another organization was like, I want to create an AI avatar that people can complain to, sort of. I don't have to listen to the complaints.

Speaker A: Well, maybe there's a place for it. But I think businesses will need to decide, like, where's their line, like, uh, what are they going to use it for and how do I want my organization to feel? Do I want people to feel like they're valued and their roles are important, or do I want them to feel like they're just training an AI to do something really basic and unimportant?

Speaker B: Yeah. And I think there's an element around, like, this technology is like, coming, and I don't think there's much you can do about it. So I think people get on the, you know, get on the train, or you, you know, go and retrain as a do. Do forestry or pottery instead, which is great. Um, but there's. Organizations can't really offer people security anymore. I don't think they have been able to for a really long time. So I don't think that the answer. You can't promise people that their job is going to be safe.

Speaker A: Right.

Speaker B: Because if you start doing that when in like a year's time, something out of your control happens, like, yeah, you just can't. You can't promise people stuff like that. What you can promise them, or promise is the wrong word, what you can offer them is like, relevancy. So, yeah, you know, your job might not be safe. I'm sorry, I can't guarantee that my job might not be safe. Nobody's job. We. Yeah, we just don't know. But I want to give you the skills and the tools and the capabilities to make you as kind of relevant, unemployable as possible. And I think people do get that. Um, I work with Mercer on their global talent trends this year, and one of the stats that I, um, like within that is, I think it's 63% of people would take a 10% pay cut. Obviously hypothetical, but, um, who kind of reskill and upskill in AI and digital tools. And that was the highest thing people would take a. Take this hypothetical pay cut for. So people kind of get it. They're worried about becoming obsolete. So what can we, as organizations, if you can't offer people security, can you offer people that kind of relevance and help them to keep themselves relevant? And obviously, you know, that part of that is on the individual to take you up on it. You can't. Like, you can lead a horse to water and whatever But I think there's an element of where we kind of redeploy, reskill, upskill, um, and giving people those kind of skills is important.

Speaker A: So we've talked about employees sabotaging AI agendas, um, implementations, whatever. Um, what about organizations that have really figured out how to get their employees on board with this shift? Have you come across any of those in your work?

Speaker B: Yeah, so there's some nice example. I did a white paper for um, business culture, Awards and connected Community. I think that was last year I believe and there was some, we had some nice examples of organizations um, there that were kind of making bit strides in it. So I think the big one is, and I said this already is, you need to recognize AI and transformation. AI transformation is a leadership and culture challenge, not a tech challenge. Um, and um, you know that's why HR needs to be, really needs to be in the room. And as well as leadership and culture challenge and, and a tech challenge, it's also an all design challenge.

Speaker A: It is, yeah.

Speaker B: All of the things I think there's kind of um, honesty and honesty about the trade offs and having this kind of adult to adult conversation rather than the parent child conversation and clarity, clearly explained logic behind why you're making decisions and transparency about impacts and you might not know what's going to change. But like I said, offering people potentially the opportunity to stay relevant as the world is changing fast I think is really valuable being principles led, ethics centered decision making, um, so kind of foundations of trust and good governance and you know, like we were talking about earlier, you know, just because you can do something doesn't mean you should. And maybe we'll come on to this later but I think removing entry level jobs is really, you know, a really strong example of you could, but should you in terms of um, you know, long time business, long term business health and also kind of societal impact. Um, and then in terms of like the kind of practical things people were doing celebrating um, failure was quite a nice one. One of the organizations talking about, you know they have the kind of, that it's quite a small organization having town halls and then people come and talk about I tried this and it didn't quite work and this is what I learned from it and kind of having those open conversations offering things like AI credits, I mean you can like reward and rank people for usage but I do think there's quite unintended consequences come to that. And I don't know if you saw like Amazon recently shut down an internal AI leaderboard because people began like Token maxing where they're basically artificially inflating AI token usage, which costs the organization loads of. By just giving it pointless tasks because they know, well, it's like any kind of reward system, right, if it's people like just gaming the system. So I'm not sure that is uh, the kind of, the smartest way might increase use. But are you incre. Are you increasing valuable use or are you just increasing use? Uh, like any, at any cost, I think. Um, so yeah, those are some examples I think. Um, and yeah, the kind of breaking down of workflows and kind of looking horizontally across organizations, I think that's like a bigger thing that requires you to think bigger. It requires you not to think horizontally. You have to kind of go east to west, so north to south across an organization. Um, and you are seeing organizations doing that. And I think that leads to all sorts of interesting questions about almost like functions collapsing into each other and where the lines of functional expertise become blurred. Um, and all of these again are big questions I'm not sure organizations are as yet prepared to answer.

Speaker A: No, I have actually seen a few um, posts on LinkedIn around should, um, HR actually report into it in order to um, help with this transformation? Um, you know, people are literally writing

Speaker B: about that right now.

Speaker A: How are you

Speaker B: not report into. It's like do the two functions merge?

Speaker A: Yeah. Oh, well that would be preferable.

Speaker B: Researching that, Researching that right now. Um, well, I don't know. You like me seeing in some organizations like that it reports into hr. So yeah, like Moderna M, um, kind of done that again, collapsed everything into resources. So it doesn't matter whether it's like human or machine. Not sure, but that's going to be the direction of travel for most organizations. I think it's more about and like this not to like sell my. Give you like what my piece is going to say before I've even written it. But basically, you know, it's kind of more about partnership. Right. And closer collaboration and understanding where expertise sits. Um, because something I have been thinking and it's probably way too deep for this conversation, but it's like what is expertise? And if AI means you can literally become an expert in anything in seconds, but it's not actual expertise because you can't question what it's giving you back. It's surface level expertise. And I think going back to the conversation we were having earlier about people encroaching onto others kind of professional skill sets, it overinflates your sense of confidence. Um, like in my AI Brain piece, which I was just reading earlier, reminding myself what I actually wrote, there was a bit in there about kind of the more you use it, the more you think it's great and you kind of start to, um, trust it more, you think less critically and you kind of swallow what it gives you. And that is, I think, the real, real danger in terms of kind of cognitive atrophy. Um, and we all talk about, you know, the kind of phrase that gets trotted out the whole time is, oh, you need the human judgment, you need critical thinking. It's like, well, how do you have those things if you haven't got a foundation of expertise? And that brings back to the junior talent piece, which is obviously an all design and HR people like, how do we, you know, if juniors aren't being trained on stuff, then how do they build their foundational knowledge that enables them to then be an effective human in or over the loop? Because you can't just do that from scratch. Like, you need to make that decision based on something. Um, and I worry that making these kind of very quick decisions through the lens of pure efficiency or pure productivity, you're gonna regret it further down the line when you realize that you might have lost any, like, level of the ability to think. To think critically.

Speaker A: Yeah, because I often think, um, the problem with using AI in our work, so we use it for certain things, but certain things we don't use it. The thing that's missing is a real appreciation of the context. So, yeah, you can put all this stuff in, you can explain things like it's an apprentice, but it doesn't know and understand the context. Um, so that, for us, that's kind of the line really. If it needs that kind of context to produce a good quality output, uh, we're not using it or we're not using it for the complete process. Um, yeah, and I just think we do actually, like, we've got some things maybe we could have looked at, um, A.I. using A.I. to, uh, manage some areas of work for us. But we've actually just brought in an apprentice because we think it's really important and that's what people buy from us. Um, applying that context to a situation, the wider understanding, the experience of what's gone before. And I think that's true in lots of, of roles and lots of organizations. Um, but yeah, it seems like an easy win at the moment, doesn't it? Just to see if Copilot can do everything or Claude or whatever you like to use.

Speaker B: Yeah, I mean, I just feel it's like, um, this again is, I would say an HR or an L and D question. And again, it's a bigger question. It's thinking about what do we gain versus what do we lose? Because, you know, there is. It's hard with the studies because, um, it's such a new technology that you can't find, like longitudinal studies what it's doing to the brain. But the studies suggest that cognitive atrophy is a thing where you're using it and then you're thinking for yourself. And therefore your cognitive abilities are, uh, kind of slightly declining. Definitely. Skill degradation is a thing, you know, a skill. You use it or you lose it. You maybe are overconfident. You think, you think you have the skill, but you actually don't have the skill because you're not using the skill. Um, but then there's also, you know, questions about, you know, I can't remember any phone numbers off by heart apart from my own anymore. And I can't read about my mum's m. You know, I couldn't, would never. Like, I can't use it. I just rely on like GPS on my phone and does that matter? Like, you know, there are, maybe there are some skills we're happy to lose. Um, one of the professors I interviewed for the, for the piece, um, that I'm talking about about how it affects your brain was saying, you know, we used to sit around campfires and like, tell, you know, the Odyssey or the Iliad to each other. And now, and like, literally when paper and pen and writing was invented, that was like, probably a moral panic about, oh, my God, well, now we're not going to be able to tell each other stories.

Speaker A: Yeah.

Speaker B: So, you know, some skills, you know, they just evolve and that's the nature of things. But I think without putting in the kind of guardrails or advice about how you use it in the best possible way. Um, you know, I think too many people use it at the start of a process and really should use it at the end of the process because you need to kind of exhaust your brain and ask, uh, all the questions, come up with all the angles, write down kind of everything that you know about a topic and then stress test it and be like, what am I missing? If I presented this to somebody who thinks like this, what would they, how would they pick holes in it? That kind of thing. Really useful. But if you're using it at the beginning and getting it to do it for you, then you're, I think, selling your brain a bit short. And that's What I would be concerned about that almost we like, we forget how to think. And I think that has long term implications for organizational sustainability and resilience and innovation, to be honest. And the human race.

Speaker A: That's been really, really insightful. Um, in closing, if a business owner was talking to you about um, bringing AI into their business, based on all the sort of research you've done, all the writing you've done on this topic, what are like the one or two things, maybe three that you think, um, that you would um, suggest they consider?

Speaker B: Um, I say I think one of the big ones would be like to start with the work and the outcomes, not the tool. So not be like, AI can do this, let's use AI, but think we have this problem and the outcome we want to get to is this or this is what the work currently looks like and how would we like to shift it? And so start with that. Not we're just going to put this tool in. So I think too many people, you know, it's very easy to take a tech first approach, but really you should take a problem first approach or an outcomes, an outcomes faced approach. That'd be one. The second one would be not to um, overlook I think the risks to um, cognition and creativity and innovation and for yourself. I think for an individual I think it's really important. Sometimes when I use it I'm like, I feel my brain cells dying so I should probably get away from it, not use it for a few days. I uh, think we all need to take responsibility for that. But I think there's an organizational aspect in how we talk to our people about the most appropriate way to use it. Um, and definitely in the education system and in how we're training young people. I think um, there's a bigger thing there as well. And then the third bit would be I really think leaders need to listen to concerns of their people. And I don't think you can just say they're Luddites or they're um, you know, they don't really understand or they can't see they're trying to protect their jobs.

Speaker A: Yeah.

Speaker B: Or they're trying to protect their jobs. I think that there's a kind of a bigger wave coming that could impact organizations in terms of kind of engagement, um, and trust. And if you don't kind of take people with you, it's not, you know, you can't kind of force. Well you can force people to use it. You can fire them if they don't use it. M. I don't think that's like a sustainable long term, long term play. So I think if the chasm between how you feel about it as a leader and how your people feel about it gets wider, you kind of, you need to bridge that through open, uh, conversation and listening, I would say.

Speaker A: Yeah, that's great advice. Thank you so much for joining me. It's been a really, really useful conversation. I'm sure many businesses and individuals will take a lot away from the conversation. Thank you.

Speaker B: Uh, thank you, Sarah. Nice to speak to.