For the last few years, most corporate conversations about AI have been about efficiency, automating tasks, and reducing headcount. That conversation is now starting to change, with more attention on what constant AI use is doing to the human capabilities organizations depend on.
Judgment sits at the center of this. It develops through years of hands-on experience, often in exactly the kind of junior work AI is now taking over. With many employers also hiring fewer entry-level people, developing the leaders of the future is becoming a serious challenge.
So how do organizations use AI to amplify human capability rather than erode it?
My guest this week is Johan Roos, Professor in Strategy at Hult International Business School and author of the new book Human Magic: Leading with Wisdom in an Age of Algorithms. In our conversation, Johan shares why the AI conversation is changing, the human capabilities that now matter most, and the practical choices employers face in developing judgment for the future.
In the interview, we discuss:
The dual mood of AI anxiety and excitement
- Moving the AI conversation beyond efficiency
- Amplification versus erosion of human capability
- The five human capabilities that matter most
- Why judgment comes from experience
- The risks of cutting entry-level hiring
- Thinking in tasks, not people
- Building an agentic workforce without cutting jobs
- Early warning indicators of talent risk
- Bionic collaboration and what the future looks like
A full transcript will appear here shortly.
Follow this podcast on Apple Podcasts.
Follow this podcast on Spotify.
Transcript
Matt Alder
Everyone agrees that hiring decisions need a human in the loop. But human judgment is built through experience, and much of that experience comes from the kind of work that AI is now taking over. So where will the next generation of leaders get their judgment from? Keep listening to find out.
Advert
Support for this podcast comes from Juicebox. If you’re a recruiter, sourcing can quickly become your full-time job. You’ve got open recs, you’ve got a deadline, and you’re spending the first two hours of your day trying to find people who might be a fit, scrolling the web, tweaking Boolean strings, sending faulty messages to get three replies. That is what Juicebox was built to fix. Juicebox is an AI recruiting platform that actually understands what you’re looking for. You describe who you’re looking for in plain English, and it surfaces candidates across 800 million profiles from GitHub, Stack Overflow, Google Scholar, and more. No Boolean, no manual filtering. And it doesn’t stop at search. Juicebox personalises outreach for each candidate automatically, pulling in from their past experience or even flagging mutual connections at your company. You can send hundreds of personalised outreach messages weekly at scale. Recruiting teams at 5,000 plus companies like Notion, Ramp, and Cursor use Juicebox to source faster, reach further and delegate the most manual parts of their sourcing work so they can focus on what makes the recruiting process feel more human. You can try Juicebox free today by going to juicebox.ai and also use the code matalder15, that’s matalder15, to get 15% off your first annual plan.
Matt Alder
Hi there, and welcome to episode 812 of Recruiting Future with me, Matt Alder. For the last few years, most corporate conversations about AI have been about efficiency, automating tasks, and reducing headcount. That conversation is now starting to change, with more attention on what constant AI use is doing to the human capabilities that organisations depend on. Judgment sits at the centre of this. It develops through years of hands-on experience, often in exactly the kind of work that AI is now taking over. With many employers also hiring fewer entry-level people, developing the leaders of the future is becoming a serious challenge. So how do organisations use AI to amplify human capability rather than eroding it? My guest this week is Johan Roos, professor in strategy at Hult International Business School and author of a new book called Human Magic, Leading with Wisdom in an Age of AI Algorithms. In our conversation, Johan shares why the AI conversation is changing, the human capabilities that now matter the most, and the practical choices employers face in developing judgment for the future.
Hi, Johan, and welcome to the podcast.
Johan Roos
Hi, Matt. Good to be here.
Matt Alder
A pleasure to have you on the show. Please, can you introduce yourself and tell everyone what you do?
Johan Roos
So I am a long-term business school professor and a business school leader and author and an advisor.
Matt Alder
Tell us more about the topics that you advise on, the books you’ve written. Give us a little bit of background.
Johan Roos
I have a bunch of books in the past, you know, a dozen or so academic books. So I’ve done anything from sort of the intellectual capital knowledge management movement in the 90s to strategic alliances, to strategy processes, to innovating strategy process, serious play. And most recently, you can say the implication of artificial intelligence on leadership. But my feeling is really strategy overlapping with leadership.
Matt Alder
So let’s dive into that AI and leadership part of things. So I know that you’ve recently spent time with some senior HR learning leaders who work for sort of big global corporations. What kind of mood are you picking up around AI? And is there anything that you think they’re getting wrong in terms of how they’re thinking about it?
Johan Roos
I often ask big audiences, and I do do a lot of talking now, about how many are kind of AI anxious, you know, hands up, and then how many are AI excited? And, of course, everybody wants to say both. So I think that picks up the mood pretty much as well. It’s a very dual relationship that most people of this level, if they are in the C-suite or even learning people development also, in a sense, there is both this excitement and this really slight fear or anxiety to it. It’s both ends of it. You find a bell curve there in different organisations.
Matt Alder
A lot of the conversation about, most of the conversation around AI is what the technology can do, the human things it can replace, those kind of things. Enough people talking, though, about the daily impact that AI use is having on humans and the capabilities that we’ve always needed to kind of rely on at work.
Johan Roos
I think less so, but thank God increasingly so as well. And this was one of the reasons I actually wrote the book that was just published, Human Magic. And because I was pushing sort of really driving an AI implementation in this global business school where I was the chief academic officer in various parts of the world. And really, everybody was exciting with the technology and all that. But on the side, I was also part of what’s called the Drucker Forum and a lot of conversation with really important CEOs, or CEOs that are CEOs of important companies, I’d rather say it that way.
But it was interesting that the conversations on the corporate side was, of course, a lot during 2023, all the way, I’d say, through half 2024 and most of 2024, was about efficiency. How do we basically get rid of people? And, you know, we can joke about that and be negative about it, but it’s also corporate reality, right? You know, you have pressure, shareholders, institutions, government, you have to make it, you know, you have to be efficient, you know, very much so.
So, of course, a lot of the talk was about that. And in the business school, how do we make this, you know, overcoming this idea of students are cheating? How do we stop this? To rather, how do we turn this to an exciting tools like the calculator, like the laptop, et cetera? It’s happened before.
But I think, in a sense, the conversation is changing. And I’m super happy about it because the timing is perfect because there’s been a lot of consultancy reports out now. There’s more literature coming out in line with what I’ve argued for quite some time, in a sense that we have to just look at it from the human side, and realise that, you know, on the big level and on micro level, our level, on individual team and organisation and even society, we’re kind of this technology is making us fork. Either we use it in a way that really grow our humanity, you can say, the human part of ourself. I’m happy to explain how I deal with that, but just think of it for a while. Or we use it as a substitute for our humanness. And that, you know, these are two paths.
I often push CEOs, CLOs, whatever, professors, whatever, and just say, hey, this is going to force you to make some serious decisions, which has to do with people, but it certainly has to do with tasks, with jobs, with people, the balance between technology and people with efficiency and effectiveness, with, you know, just rushing ahead, you know, doing the same versus imagination and all that.
So there’s a lot of exciting things. And if you change the focus or unit of analysis from the technology to us humans, I think we will have, we’re having much more interesting conversations. Much more interesting. And it’s going there.
Matt Alder
Your book’s called Human Magic. Talk a little bit about that humanness. What is it that’s really important? And how does it fit with AI?
Johan Roos
It does, in a sense. And it’s not, I’m a very sort of tech optimistic person. Believe me. It’s like, I love this stuff. I have dozens of agents working for me, and I’ve done voice assistants and all kinds of cool stuff from January 2023. I’m all for this, but I discovered very quickly that the risk in this is that we forget what makes us human.
I’m not in therapy, I’m not in psychology, I’m in leadership and strategy and management. My focus is on us as professionals, professionals at work. Though there’s a lot of discussion about what our kids and so forth do.
So that’s it. So I took it all myself to, hey, just basically remind people in a common sense way and say, hey, you have a choice here. We have a choice here, how we do this with this fork. And it’s your choice. It’s really your choice what you’re going to do with it.
So I talked about this humanness or humanity in these basic human capability terms. So, the five Cs, you know them, it’s curiosity, it’s creativity, it’s critical thinking, it is the way we communicate, and also how we collaborate. For me, that is super essential who we are as, and I will say, as professionals, of course. Private individuals as well. But as professionals, these are the tools, the skills, the capabilities we use at work.
And on top of those, or underlying those, is this notion of expertise, eventually judgment, which has been discussed throughout history as wisdom, or to quote from Aristotle and Thomas Aquinas, et cetera, practical wisdom, which is a concept you find in different languages, in different cultures, basically the same. The whole idea of being able to answer what really matters here, what’s really right to do, what resonates in the community I’m a part of, and what will work when I get going on this. These are essential questions that I’m super worried if people leave those questions to AI to answer. Because that is to me like, if you do that, you really become what I call in the book an AI concierge, which is actually a quote from a CEO. Instead of being sort of a real good citizen, a professional citizen, you become sort of an algorithmic citizen.
So that is how I’ve taken it on to kind of just say, hey, this is what it is. Warn about it. But here’s how you do it to do the amplification route. And here’s how you kind of hold on to your human magic.
Matt Alder
I want to dive into that and the judgment aspects of this in a second. But before we do, though, just revisiting the five Cs, I mean, traditionally, they’ve been areas that organisations haven’t really sort of invested in, in their employees. What are the implications around that for employers that these are the skills that are now absolutely critical?
Johan Roos
You could say that we all, individual and our employers, kind of do invest in this, but not explicitly. You know, people talk about, hey, you should be creative, blah, blah, blah, and all that, but maybe perhaps in certain fields like accountant, not too creative, please. But, you know, this idea that people talk about, it’s part of all leadership programmes, but hidden, I think, hidden. You know, communication is one thing that is more explicit, like here’s how we communicate and so forth. And teamwork, collaboration, all that. So it’s there, but it’s never, I think, been brought to the proper forefront. It’s always there in the oft, explicit or somewhat explicit, a little bit more implicit. So what I wanted to do was to help bring this into the forefront.
In fact, I will argue, and I just came back from a big higher education conference in Seville, that this should perhaps be the core of what we teach people. It’s almost like back to the Renaissance idea of a liberal arts kind of a liberal education, not not anything else but the notion of liberal in the sense that it’s helping you free yourself from dogma, you know, you should be a human being that can assess, you know, be critical, be creative and all that to help you not fall into the dogma at that time of the king or the church but really to be a good citizen. So it’s nothing unusual or new. It’s just a reminder of this.
And I think what I’m trying to do, and I actually get some traction on this, is to help people think through, wow, okay, good reminder, what do we do to do this now in this enormous colonisation of AI agents throughout the organisations? And it’s going to be a challenge, but there are ways of doing it, of course.
Matt Alder
Moving forward from that to the judgment aspects that you talked about, there’s just such a, you know, everyone talks about human in the loop. We need human in the loop to do this. We need human in the loop to do this. Obviously, in a lot of cases, that is around judgment. So I’m thinking sort of particularly of recruiting at the moment where, you know, final decisions made by humans, all that sort of stuff. How do we protect the ability for people to have judgment if more and more things are getting outsourced to AI? How do people have that experience too?
Johan Roos
This is a real concern now. I think it should be an even more of a concern to people. But some people are, you can say, wakening up to this challenge. Challenge, how do we get our judgment?
How do we possibly become better at making decisions, Matt? Well, through our experience, our experience for good and for bad. Actually, the more failures and stuff we have, the more we can learn and move on. And it was sort of adjust our fine-tuning our judgment. If we now, you can say, outsource some of the, what should we call it, grunt work, the early junior type of work that you and I have been doing. Draft reports, doing this, experimenting with that, getting slapped on the finger, getting feedback and all that, that will shape our sense of, oh, what is right and what’s wrong, what’s good and what’s bad, what’s beautiful and not, all of those serious considerations that are fundamental. If we outsource that to the machine, it is, I think, shooting ourselves a little bit in the foot. It is taking away the opportunity.
And if it’s something I’m a little bit concerned with, is the data about new hiring and companies hiring less and less junior people, and why graduate students have a little bit more of a difficult time now, especially in broader social science fields. But so I’m urging everybody, like, don’t stop hiring. You know, keep on fresh blood, because otherwise, what kind of judgment will the next generation of leaders have? Whenever they have to make a decision, do they have to cling on to hold on to their LLM, to their AI bot, to their AI agent or their swarm of agents? Or actually, could they have something we still call a hunch, an intuition for what’s right and wrong?
So I think this is super serious, a societal challenge, Matt, that we think we really need to take care of. And having spent a lot of time in higher education, I do take this very seriously, and I talk to a lot of people in the industry about this all the time. How do we manage, in a sense, this balance between this?
A first advice I give to managers and leaders is actually change your unit of analysis from people to tasks. And it sounds so simple and obvious, but, you know, it makes a heck of a difference if you start to think of being, you know, more efficient and you look at people, or if you say, we’re going to be more efficient at using these tasks. So I have some good stories on that. But that’s very helpful, actually, just to change that mindset.
Matt Alder
I think you’re absolutely right, because obviously jobs are made up of tasks, and you see people currently laying off lots of people because of AI. But those people do also have tasks that AI can’t do. So it doesn’t really make sense to do it from that perspective, does it?
Johan Roos
Yes, yes, yes, yes, absolutely. And this is the beauty. I have one good example. I can’t mention it, but a really big business-to-business platform type of company, global, with, God knows, 23,000 employees, $80 billion, whatever. I just talked to them, and they did something really smart, which was to say, hey, we want to become an agentic workforce in one year. And everybody said, well, what does that mean? You know, the CEO said that, and he didn’t define it. But then he says, well, the way to do it is we all should develop two agents per person, 50,000 AI agents across all function, from accounting to strategy. In six months.
And then we need to train people on levels, like the basic agents. Everybody should be able to do basic agents, such as doing an agent that reads your email and draft a response. Draft, not send it. Draft. And then there are levels to that more advanced. And they set up competitions and awards and everything. And they say, this is not about taking away people. This is about changing jobs. And not just becoming more efficient at what we do across the board, but also innovate.
And lo and behold, six months later, 54,000 agents had been developed across the world. And they said there was a bias in the sense, like a lot of the young people and some of the interns were the most prolific at this, of course, because they’re the generation. But everybody had to go through this. I thought it was fantastic. People were excited. Nobody got fired because of this.
So in a sense, it was a very, very clear signal. We’re looking at tasks so that we can work smarter and in all kinds of ways, smarter and more efficient. This is not about removing people from, you know, somebody got fired or was from another reason. But that was a super, I liked that a lot. I liked, that was a good story.
Now, of course, they said, how on earth do we do this in reality when we implement this? And what do we do when we have AI agents as teammates, which is something I talk about in the book. I call it bionic collaboration. That’s an essential skill now for all of us.
Matt Alder
And I want to talk a bit more about people coming into the workforce, because you were talking there about people shouldn’t be cutting junior roles. But how do those junior people get experience? Because I’m kind of looking back at my career, where my judgment is, and that’s come from many decades of doing work, and particularly in my earlier career, work that AI would now do. So how can people coming into the workforce get the experience to have that judgment?
Johan Roos
And I think a big, you know, we learn on the job, right? Okay. On the job, learning is perhaps the most important one ever. At universities, at grad school, whatever, we learn knowledge, but we also learn our social skills. You know, there’s a whole range of this. But if you think of it, it’s like, I think higher education needs to reinvent itself a little bit more and a little bit faster than it’s doing now where, you know, universities were built to last not to change.
So this idea, but I think I’ve always been a proponent for two things which are, I don’t see a lot of. One is thinking of higher education a little bit more as a vocational in its positive sense, not in a derogatory sense, but really like why don’t we take a week at uni and then a week at work and why don’t we take it two weeks there and a week over, so kind of intertwine this work thing. Some countries do this, and they kind of build sort of the apprenticeship model from that. And this is typically the case in, of course, vocational training. You know, you become an electrician, a plumber, or a carpenter, or whatever. It’s back and forth, back and forth, back and forth. That is a very good model, I think, even for higher education. You know, you need to be able to study physics, but you could actually do something where you see engineering happening. And certainly in the business studies, which is 20% of all higher education, I’ve always asked for that.
The second thing I’ve really pushed but never succeeded with is to say, hey, guys, lifelong learning is a nice propagandistic statement, but what do we do with people that are about to retire? Why don’t we bring them in as the coaches in the apprenticeship sense of the upper phase where you can come down in and start to do something and bring in all of this enormous experience that otherwise would be on the beach in Florida or play golf or whatever, but really come in and help more. It’s just that unis are not built for this, right? Companies are not built for it.
But again, I think we have a lot to do here to ensure that the model is not just staying, but rejuvenated in a world where the machines will do a lot of the tasks. Everything that can be automated will be automated. So we need to hold on to this magic and the creation of this human magic that is in these experience-based learning.
Matt Alder
There is a huge amount of disruption to come and an awful lot of work to do. Just for CHROs, heads of talent, people sort of organising, you know, work as it were within their company, what would your advice be in terms of what do they need to do and think of to kind of get this right for the future right now?
Johan Roos
Whatever I say is within the context of corporate realities, right? You know, it is the companies are not voluntary organisations, right? There is a job to do and there’s profit to deliver and there’s investments to be made and there’s like, you know, efficiency to be gained and so forth.
So, having said that, I think there’s something on the highest level, which is about being a role model to do this. Whenever I’ve been a leader, and I’ve been a leader of two organisations and part of the team of another two, three, actually, and then part of the leadership team of another two, it’s like if you’re sitting behind and watching and telling people what to do, that’s one model. The other thing is like, well, you’ve got to get your hands dirty by seeing this. The CEO I was just telling you about, about these 50,000 things, out there on the barricades in a sense, doing this stuff, showing. I think now with AI tools, I think all leaders have to have a fundamental knowledge of what they talk. Otherwise, what can you say if you don’t know what you’re talking about? You have to understand what is an agent, what can they do, what is a swarm of agents, what’s happening, not just ChatGPT in the evenings. It’s not about that. It’s really about seeing the potential.
I always tell CEOs, for example, and CHROs, I say, you got to get your hands dirty. You have to show that you comprehend this, ask your kids or whatever, grandkids to help you, but it’s got to be you have to do something. I think that’s important.
The other one, I think, is really about basic change management. This is one change automatic. Everybody has been through change management. We know that. It’s nice to say, hey, change, and everybody will nod, but everybody will say, after you. So it’s about making change exciting, like the case I just told you about. Exciting, not threatening, which is typically the case. Actually, people say they like change, but nobody wants to be changed. And we have to kind of respect that as well.
And I also think it’s all the way down to, what should we call it, metrics. Again, what are your metrics? What are your metrics in terms of measuring progress? If it’s efficiency metrics, well, then you will look for one thing. People will tend to do what you measure them on and you incentivise them on. Do you really measure the effectiveness of erosion of your capability, or are you kind of innovating new metrics for amplification?
I’m working actually on the measurement model on this. I’m an indicator model of the five Cs and practical wisdom with a super cool American not-for-profit, to sort of building this instrument of a little bit of alternative measures to do this.
How can you measure talent risks? How can you measure these risks? How do you know when the curiosity of individuals and teams start to go down or actually amplify? How do you know? It’s pretty obvious in communication. On communication, people start to write like an LLM because they’ve copied and pasted something from Claude or Perplexity or whatever. And but how do you then collaborate? There’s a lot of interesting things that you can measure and you can indicate, erosion or amplification, and I think this is super important to rethink your measurement models across the board to just try to not to be perfectly right but roughly, roughly helpfully and say, hey, here’s some early warning indicators. So just some examples.
Matt Alder
Final question. Just give us a little bit of an insight into your sort of view of the future, because you mentioned sort of humans and agents working together. What’s work going to look like in two or three years’ time, do you think?
Johan Roos
Well, two or three years’ time, yes, yes. You know, I’m not a vivid reader of science fiction, but I’ve gone through some of the classics. And it’s interesting that in science fiction, you know, if you think of Philip K. Dick, if you think of Isaac Asimov, some of these things are also in pop culture as very popular movies, right? I, Robot, 2001, and all that, even Blade Runner. They already ask some really good questions, even the Bicentennial Man, what happens? If we are sort of working with robots, what happens if robots come into our family? What happens if robots live forever and we die? You know, there’s a lot of these questions that have already been discussed in, I say, high quality science fiction literature.
So, I mean, if anyone wants to sort of increase their mindset a little bit on this, look at some of this stuff, because they ask very good questions. What happens if you long live forever in a society? How does that impact society and so forth?
In the shorter run, of course, we have a tendency to be overexcited about everything new. That’s our humanness. We exaggerate the short-term impact, but we actually miss out. We underestimate often on the longer-term impact. You could say in two to three years, I think we will see amazing new AI tools. Look at the data. Look at what they’re announcing. The tech bros and those guys are announcing every week now, and it’s like doubling, tripling every four months. This is beyond Moore’s law.
It’s extraordinary what you can do. We may drain our planet of energy very quickly here, unless we get these micro nuclear power plants, et cetera. But it’s going to be a heck of a ride now over the next couple of years. A heck of a ride. It’s going to be super competitive nationally and internationally between the geopolitical blocks forming.
So I just think like fasten your seatbelt now because it’s going to be a heck of a journey and I think, I think we all, Matt, have to work really hard to stay, to not fall behind, to not fall behind. Because when the environment are moving so quickly, doing the same thing is not a good idea. So we have to work really, really hard the next couple of years.
Will it stabilise? Well, societies tend to stabilise. And then there are these dynamic equilibriums. And suddenly, there’s an earthquake of some kind. There’s a war. There’s a new president. There’s a this or there’s a that, a new invention. That’s how we work. We think it’s stable, but it’s not. Change comes not linearly, but very exponentially and sudden. It’s called complex adaptive systems in theory. This is how it works. So I think it’s going to be a heck of an interesting place now. I don’t have the answer to who’s going to win and all that. That’s not important. But the thing is keep an eye on it, work on it, move ahead, move, move, go with the flow and try to learn as much as you can. Otherwise, you will, I’m afraid, fall behind and the world will be the do’s and the don’ts, the have and the have nots, and the will do and the won’t, which is not a good picture.
But I am an optimist. I’m an optimist, but there is a darker side, which I don’t want to talk about. But there is like, and I think most professionals will come to realise that they have to hold on to their human magic. Otherwise, what value do they create? What value do you create if you only can ask for a prompt and then do the prompt and then put it together in the report? That’s not the point. The point is to use these amazing tools to amplify your curiosity, to become more creative, to become more critical in your thinking, to ask the right kind of question, to actually communicate still using your embodied capacity, where you see the micro expressions in people’s faces. We can’t smell each other now, Matt, but this idea of being present matters. And then, in a sense, this notion of collaborating. We’re really smart in collaborating with the machine. And again, the literature has covered this many years ago.
Matt Alder
That’s my summer reading list sorted out, definitely. Good. Johan, thank you very much for talking to me.
My thanks to Johan. You can follow this podcast on Apple Podcasts, on Spotify, or wherever you listen to your podcasts. You can search all the past episodes at recruitingfuture.com. On that site, you can also sign up for our weekly newsletter, Recruiting Future Feast, and get the inside track on everything that’s coming up on the show. Thanks very much for listening. I’ll be back next time. And I hope you’ll join me.
