In this episode of The Parlor Room Presents: Hello AI, host and Harvard Business School Online Creative Director Chris Linnane speaks with HBS Professor Linda Hill about why AI adoption depends on more than technology. Drawing on her research, Hill explains why leaders must rethink the way work gets done, teams collaborate, and organizations build the trust and capabilities to use AI effectively—and why the most successful companies treat AI as an opportunity to redesign work, not just automate it.
Resources
Learn more from Hill in her course Leading in the Digital World, which is part of HBS Online's Credential of Leadership, Impact, and Management in Business (CLIMB) and the Credential of Digital Innovation and Strategy
Catch up on the previous Parlor Room episode, "Linda Hill on Leading Change and the Paradoxes of Management"
Check out Hill's book mentioned in this episode, Collective Genius
You can also watch The Parlor Room Presents: Hello AI on YouTube.
Transcript
Editor's Note: The following was prepared by a machine algorithm and may not perfectly reflect the interview's audio file.
Linda Hill: So if you just take what you're already doing and just use AI to help you with it, it's going to be ineffective.
Chris Linnane: Linda Hill is a professor at Harvard Business School.
Linda Hill: What will my workforce actually look like? First off, what will the work look like, and what will my workforce look like? These are unknowns.
Chris Linnane: One of the top 10 management thinkers in the world. Her research helps organizations lead through transformation.
Linda Hill: One organization, which got rid of about two-thirds of its customer service people.
Chris Linnane: Wow.
Linda Hill: They had to hire almost all of them back.
Chris Linnane: She has seen what happens when companies get it wrong.
Linda Hill: Eighty percent of them said they thought it would be complementary. Within a year, 53 percent indeed more of them were beginning to think it was going to be a substitute.
Chris Linnane: Kind of fosters this "Ready, fire, aim" mentality.
Linda Hill: Yes.
Chris Linnane: Her research on innovation and transformation is exactly what leaders need right now.
Linda Hill: Because we certainly find that unless the C-suite is engaged in a digital transformation, it moves very slowly.
Chris Linnane: Welcome to The Parlor Room Presents: Hello AI.
Chris Linnane: Before we start, I want to take a quick moment to thank you for listening to this season of The Parlor Room. If you're watching on video, you can see that I'm in Klarman Hall right now in the Klarman Auditorium. Please follow on whatever platform you're using to watch or listen to the show. Thank you, and I hope you enjoy this episode.
So we're early days here, but when did AI for leaders become something they didn't just use, but they realized they had to actually lead as well?
Linda Hill: I think leaders recognized very early on that there was something very different about gen AI. One thing that worried many of them is they didn't understand what it was or what the possibilities were for it, and they knew it was very early on, and they needed to get on it, so they weren't using it a whole lot. I think many of them use it now in some form or fashion, but they weren't using it a whole lot.
But when they first heard about it, they became nervous. For the last five years, we've been collecting data, doing round tables, and doing surveys of leaders. We've surveyed over 8,000 leaders at this point, and really engaged another 300 in round tables. And when gen AI got announced, the survey data was very interesting. The round tables, what is this? They were talking to each other about it. How are you going to use it? Who knows? Oh my goodness. So I know they knew it was something different and you had to learn about it, and they were going to have to know something about it in a very different way.
Chris Linnane: And when you started talking to them, just overall, was the mood excitement or fear, do you think?
Linda Hill: What was their mood? The mood was, for most of them, it was a worry. They were already feeling pretty overwhelmed. And we have to remember, of course, they had just come out of COVID, all of that, so all of that change was happening, and now there's this, this gen AI thing. And everyone says it's going to make a difference, it's going to change the world. I never really completely got what the difference was between machine learning and AI, and now we have this gen AI, and what's this all about? So I would say they were curious about it, but they were already tired, and they were already thinking about all the change that they had to deal with already, so mostly I would say, "Oh my goodness, this is something else I better know about."
Chris Linnane: Yeah. The best time to get bad news is when you're already dealing with bad news, I guess.
Linda Hill: I suppose that could be it. So I don't think they were; they didn't have a sense that this was going to be a disaster or anything. It's just like, "This is another thing that has to be dealt with. And frankly, if I'm not a techie, I don't really understand this and what its capabilities are, and I need to get on top of it because there's a lot happening here."
Now, one of the things we see is that many people have come to think of it as a bit of a hype, and it's not really ready for prime time for some of the use cases they hoped to use it for. However, they know you need to be practicing and experimenting with it now.
Chris Linnane: Now that hype, that's an interesting word because I feel that too. I use the terminology or the word, the phrase magic trick a lot, where it feels like at first you see something, and you're like, "Oh my gosh." Then you realize, "Okay, well, I was just dazzled by the fact that it happened, but can I really apply it to anything yet?" So was it pessimistic about how are we actually going to use it impressive, but how are we going to use it?
Linda Hill: I think really, when it first came out, and early in 2023, we happened to do some round tables. I think they were all just trying to understand what it really was. And I think the other thing that I should have said earlier is that they were already going through digital transformation, and a lot of those transformations were not going as well as people anticipated. So despite all the investment they were already making in technology and data, digital technology and data, they weren't really getting the payoff. So then this was just another piece of that puzzle. And so that was a worry for them.
But on the other hand, some were quite excited about it. And we obviously have been setting CEOs who actually run some of the organizations that are developing these tools, but mostly they were already, they were tired, the digital transformation is not going as well as we had anticipated, and now there's this other thing. And there's one very specific piece of it that worried them.
In our own data with the 8,000 plus, it turns out that it's not easy to get people to embrace data in doing their decision making, not at all. Again, when they heard about this new technology, and it's not clear, it's rather ambiguous, it's like, "Oh my goodness, I haven't even gotten them to use some of the basics that we already have in place. And now there's something that's even more complex, and how am I going to get them to use it?" Because all of this, it's all very costly. And there was a lot of cost pressure on organizations even in 2023, and that just continues to be true.
Chris Linnane: Now, if we can, I'd like to go a little bit deeper on that. So does AI break the way leaders learn to manage digital change?
Linda Hill: It doesn't break the way they've learned to lead digital change. In fact, they have a lot more to learn about that. Again, you're adding this other tool that they don't really understand to something that really wasn't going as well as most of them had hoped in the first place. It doesn't break the way. It does indeed over time as you begin to use it and you think about how you're going to structure your organization, and whatever is going to happen, they are going to have to make more changes.
So what I wanted to say to put this into context is that what we find is that those leaders who are making the most progress and getting people to actually embrace and use these tools to make a difference and create value, those leaders have reframed how they think about digital transformation. And they've said it really is about us getting people to work differently, to use the digital tools we already have, the data we already have.
Those leaders who are already beginning to do that anyway are probably not going to have to change so much the way they're going about doing the digital transformation, and I will talk a little bit about that. But for those who are still struggling with trying to get that right, this just made it more complex. And it turns out that many, many leaders, and they discover it over time, and they tell us this, these are senior people, they really thought it was mostly about technology, and then they discovered it was more about people.
We collect data on what you focus your time and attention on as a leader when you're doing a digital transformation. It turns out that the employee experience, most of them weren't very focused on. In fact, I think it was about 12 percent who really focused on the employee experience as they tried to do a transformation. Now, when you do focus on the employee experience, you're much more likely to have some success in getting them to embrace those tools. So when you ask me, did it break it? No, it just meant they needed to go back and look at how they had been doing things, and it was more urgent that they figure out that this is really about the employee and about the culture and the way we do our work. And so that means when you're trying to do cultural transformations, as we all know, that never happens fast in a large organization. And we all know that the senior people need to be deeply engaged.
So I think it just meant they had more to pay attention to because we certainly find that unless the C-suite is engaged in a digital transformation, it moves very slowly because, really, it is about a new way of working, which goes back to your culture and capabilities. And so it is about figuring out how are people going to use these new tools to enable them to do work differently and come up with really more innovative services, and products, and experiences for our customers. So that I'm very surprised and pleasantly surprised, and I don't think it's just because I'm the common denominator here, but I work with two separate teams of collaborators.
One set that has been working with me on what it takes to build organizations that can innovate time and again. The broad question is, we've been doing that work for almost 20-some years now, and the other team is very much focused on how to do digital transformation. The results of what innovation and the digital, they actually, they come together. They're very similar, the issues, because really to use the tools, the digital tools, gen AI, et cetera, you in fact need to get people to be able to collaborate, to experiment and learn together. And that's what you need to do for any kind of innovation.
So it's amazing to me how the research has become. And a lot of the innovations these days are about using technology. So maybe that's not so surprising, but we all know, and this is what people tell us all the time in all of our surveys, if you want to build an organization that can flourish now, given all the technologies, you have to have people in the organization who will use data to make decisions, you have to be able to collaborate. You have to have cross-functional collaboration, you have to be customer-focused, you have to have people who are continuously willing to learn, and people have to be comfortable with change and ambiguity. That's what they tell us.
We have gotten the same list every year since 2022. As a leader, I need to build an organization that has those characteristics. And those characteristics are very tied to what it takes to get people to embrace AI. Now I'm starting a project, actually, I've already started it, but I'm going to collect the first data specifically on how you lead your organization, so that you have human and agentic AI collaboration.
Chris Linnane: Tricky.
Linda Hill: I think that I will learn lots of new things and there will be, there's something else you need to think about before you're going to be able to do that besides what we're seeing in our other studies. But fundamentally, we're still trying to get organizations are still too siloed, right? Organizations, people still aren't so good at experimenting and learning in very efficient and effective ways across organizations.
Linda Hill: And that's what's required for the implementation of gen AI. The other piece, and our new book, Genius at Scale, talks a lot about this.
Chris Linnane: I have the book. Oh, I have the book. I have the book right here.
Linda Hill: Oh my goodness.
Chris Linnane: Yep, so here it is. Here it is. I'd love to hear more, but tell me about the book too, so finish your answer, and then let's talk about the book.
Linda Hill: What I wanted to talk about in the book, and just one quick piece of the book, in the book we talk about three roles, three leadership roles that have to be played if you want to be able to not only generate innovative solutions, but actually scale them. Make sure that they get launched, commercialized, they have the impact you want. And it turns out the three roles are the architect, the bridger, and the catalyst.
Well, the bridger, these are individuals who know how to work across organizational boundaries. That is very important when it comes to digital transformation, and one of the things we see in organizations that are trying desperately to embrace gen AI is that they don't have enough bridgers. That is people who know how to work across organizational boundaries because you need people who understand the business well enough to make sure that whatever you're designing, however you're using the gen AI, whatever those use cases are, to be able to understand if these are the right use cases and what the output is creating, does it really fit what we need to have happen?
Since gen AI doesn't have tacit knowledge, it doesn't have contextual knowledge; you need to have people around to understand both in the design and the deployment of it to make sure it's working the way you want it to work. Those people are bridgers. And one of the things I keep hearing from so many senior executives is, in my language, we don't have enough bridgers. We don't have people who can work across "business and technology." If you don't have that, hard to design the right solutions for your customers, and very hard to deploy.
Chris Linnane: Yeah. Now, I don't want to go away from bridgers just yet because that sounds like such a valuable position for people to put themselves in, and everyone's nervous about the future of employment and different things like that. Bridger is someone who can play all the positions, help out, and translate. How can people prepare themselves to be good bridgers?
Linda Hill: Well, I don't know that you would know anything about my early career, but where I started is I studied how do stars learn to lead? And in fact, in doing that work on how stars learn to lead, I studied, for instance, a number of people who were engineers, and engineers who had to then become managers or leaders of other engineers. There are lots of reasons why stars have more trouble learning to lead, so let's just set that aside for the moment.
Chris Linnane: Yes, okay. Interesting.
Linda Hill: And a lot of that research was started at MIT, where they have lots of very smart, star-like people who, for whatever reason, have trouble leading. So that is actually where I started my career, studying those kinds of questions. When you look at how you are going to prepare people to be bridgers who are tech-savvy enough and know enough about the business to be able to help with this design and deployment of these tools, it's a tricky question for a range of reasons. I actually spoke to a company yesterday that was mostly engineers about why they might be having trouble making sure they're using these tools that are very expensive to use in the most efficient and effective ways.
One of the things you need to do is you need to help people be more T-shaped as it's described. That is very deep in some area of expertise, but broad enough across the top that they can use that expertise in collaborating with others with different expertise. What you see to develop that, you really have to have what is referred to as a zigzag career. You do a little of this, little of that. When we look at the ones who've become important bridgers in these companies, they've often been in marketing for a while.
Maybe they've done some sales, maybe they have computer science backgrounds, and maybe they have been in IT in some form or fashion, but you see that they're pretty broad. They've had a range of experiences. So many companies don't let people develop that way; they're pretty, you have your expertise, and you stay in your box or your lane.
That doesn't work when you're trying to create bridgers because what we know is based on what we get to do. We don't learn in school how to do these jobs; we learn on the job. Unless I get an opportunity to work across differences, I don't develop the capacity to embrace and negotiate differences or figure out how to manage the conflict that is inevitable when you start working with people who come from another world, if you will. And a lot of the bridgers that we're looking at now for gen AI, they are actually having to work with people from different organizations. Think about Microsoft working with OpenAI to get Copilot. It's not just about the top doing the contract to say, "What's the relationship going to be? IP, et cetera." You literally need people all down the line who can actually work across that organizational boundary, this big Microsoft with more startup-like OpenAI.
Linda Hill: More startup-like OpenAI, you have to have people who know how to translate across and integrate so that you actually are co-creating a solution together. So again, I think it's how do you prepare yourself? You realize that you need to take charge of your development and make sure you get some of those zigzag experiences. May not be in the company. It may be that it'll be something you'll do outside the company because you're not getting a chance to be exposed to learn those possibilities, those kinds of capabilities. So you need to figure out how to do it.
Chris Linnane: Now you've solved the problem for me just now, as you're speaking, in that you always hear the Jack or Jill of all trades, master of none. But I can rethink about it now as the Jack and Jill of all trades, master of bridging.
Linda Hill: But you do have to, yes, master of bridging, but you do need to have some expertise. So one of the things I frequently have said to the MBAs, to develop expertise takes a long time because it's not really the judgment required. And judgment comes from experience. So you may go to school, but then, as soon as you go to work, we all find this experience, all of us. Guess what? They didn't teach us the things we really need to know to be able to do this well.
And you also, again, when you're working with gen AI or gentech AI too, the issue is the human is at this point anyway. You're bringing the judgment to the equation, the solution, right? Because these are just mathematics. These are equations that are running, right? Someone's all of that. But to figure out the use case and whether it's how to design it in a way that it will solve the problem that you want it to solve, use case and/or to look back and say, well, did it really do it given that we know the technology is still evolving, you have to have pretty good judgment about what really matters.
And one of the things that executives are telling me who are deploying all of these different tools is you have to keep reminding people that it isn't human because one of the things that sort of tricks you about it is it looks like it's human. It's speaking in natural language. And so I was reading a study, and I don't have the statistics in front of me. It turns out that people, if you tell people this solution as the result of machine learning, some percentage trust the solution, others don't. A large percentage don't. When gen AI does it, a larger percentage trusts it.
Chris Linnane: Why is that? Do we know?
Linda Hill: Well, because it's in natural language.
Chris Linnane: Okay.
Linda Hill: And you type the way. And it is not accurate, though, because in fact, I mean, there are two different sets of tools. But in fact, the machine learning answers and these people who do these studies are actually much more accurate than what's coming out of whatever a large language model has produced for you. It's just that it looks...we anthropomorphize. We sort of say, "Oh, no, not true." So you actually need to have people who have good judgment to understand that this... Even I'm sure you've gotten some answers to some things, and you think they're true and they're not at all true, but it answers you in a way that makes you think, oh, this looks pretty authoritative, and it went through a computer.
Chris Linnane: Yeah. And I've talked about this before is it makes me think I'm a genius at the same time because every idea I came is outstanding, every idea. And then it takes me a little while to realize that was not a very good A
Linda Hill: Total hallucination, right?
Chris Linnane: Yeah.
Linda Hill: Those facts, that place doesn't exist. Literally, whatever it is. So I think that one has to...so that is one of the reasons why I think with these tools, you do need to educate people. They need to have AI literacy to understand what it is, how those solutions are being produced. And I'm not negative about this. I'm positive about it. But it's funny, because I have a PhD and had to take lots of statistics, I understand probability theory, but most people don't. And a lot of people have trouble with statistics. A lot of people, right?
Chris Linnane: Right here.
Linda Hill: Really smart people. And that's fundamentally what you need. You kind of need to understand how those probabilities work and why it's happening the way it does. So one thing that I've been doing...now, one of my responsibilities is to figure out how do we help leaders know what they need to know so that they can, in fact, use these tools appropriately and build organizations where people are using them appropriately. So I have been invited to...the head of the AI department of MIT has sort of a monthly salon at his home, and the doctoral students and other people just present what they're working on.
And I would say at least half of what they talk about, I have no clue what they are talking about. But I'm hearing the language, and I'm hanging out with people who are doing these things, who are at the cutting edge. But it's interesting because I've made this commitment to go there and do this because if I don't have a bit of a sense myself how this all works, I don't know how I'm going to work with...and we're going to be using it here. We do use it here at the school, and am I comfortable with it? Yeah, I need to make my own judgments about that, but how am I going to help other executives understand about this? So it's not easy. At least I'm a professor, and I'm used to being with students and all of that and feeling stupid a lot of the time. But I think a lot of rather senior people, the idea that you would sit with a number of people who are...I'm having trouble getting to the ground.
They all sit on the ground. I got to get down. I got to get up. And I'm with a bunch of young people who are really extremely knowledgeable about the technology of all of this. So interestingly enough, they asked me if I wanted to come speak about the leadership parts of these things, the organizational pieces of the puzzle. And I say this to say that really one of the things we're actually seeing is that those leaders, those senior executives who are figuring out how to leverage their digital natives also seem to be making more progress. Now, I don't know the cause and effect here because a lot of senior leaders do tell us, "I don't understand this generation, Gen Z, and I'm not so comfortable with them," whatever, but I need to get to know them. And I think probably since they're so eager to use all of these tools, I can figure out how to have them be some of my champions, to be champions for me in the company.
So we do have lots of stories of very senior people figuring out how to interact with...they don't interact anyway, very senior people often with very junior people or who are young. Now, the young people who they interact with, and there are different things. We've written an article about this, but one of the things that's interesting about it is that there's some reverse mentoring in a way that goes on, but part of it is that they're a cut-and-try generation anyway, so they're just going to cut and try. They don't plan, they just do it. And it's really hard to tell them, "Don't use these tools." They are very clever at figuring out how to use them.
So what these leaders are finding is they actually need to educate their digital natives about the business. They need to educate, give them some big picture sense of it, help explain the why behind any kind of restrictions or compliance or whatever, really talk with them about why those exist, so it doesn't just feel like bureaucracy. And also, some are even engaging them very actively in helping to not only design the tools, but to come up with the governance structures as a way of both sort of understanding how these people who seem rather alien to me, how do they think about what these tools are?
And then with that understanding, if I don't understand that and I just design governance systems, et cetera, they may or may not really be getting at what they're prone to do, that's going to make us really crazy or hurt the business. And frankly, they know a lot more about it than I do in some ways about the actual technology. So we do see this sort of reverse mentoring, specifically with gen AI happening.
Chris Linnane: Yeah, I imagine. It's interesting. That's a challenge I hadn't really thought about in that way, but the stakes and the time are usually so low in the experimenting of the gen AI. It kind of fosters this "Ready, fire, aim mentality," which is really, in a way, could be so useful.
Linda Hill: Yes.
Chris Linnane: But that's where the leader has to be like the experimenting, all that's great, but we can't just experiment endlessly, and we need to think about the rules and how we're going to do things. That's an interesting challenge for them.
Linda Hill: And also how you're going to, really, how you want to design things. So I was in Japan, and I was on a panel on innovation, and there was a man sitting next to me who seemed younger, a younger man; he was the youngest person on the panel. And he began to talk about teleportation and his new startup in teleportation. Now, I love Star Trek. I love Star Wars, all of that. So I thought, did he say teleportation?
Chris Linnane: And he did?
Linda Hill: And he did say teleportation. So when the break came, I think I was one of the few people who kind of walked over to chat with him. It was a very nice reception, but I wanted to find out more about this. And so I, "Did you say teleportation?" And he said, "Yes, that's my startup." Now I want to tell you that the startup is really doing very well, and what it's about is really teleportation without the mass because when he went to talk to quantum physicists to discover when will we be able to transport the body, they said, "It will probably be another 60 years." So that's a little bit long. He said, "Well, what can we do before that?" "Well, let's transport human consciousness. Let's transport the five senses."
Chris Linnane: Oh my gosh.
Linda Hill: And what really drove him about this was the fact that he didn't understand why resources couldn't be more distributed in a more democratic, equitable way around the world. So he's a marvelous human being, Akira is. So his partner, Kevin, they started this business. And what it is about, they're in Japan, as I said, and Japan is one of the leaders in robotics. And this is a robot where you put the human in the robot. So it's more than what a human in the loop. It's actually human at the helm.
Chris Linnane: Oh, interesting.
Linda Hill: Because you're in it. So you can be sitting wherever you are. I can be here, and there is one of these avatar robots, their general purpose. You rent them like you would rent an Uber for some period of time to do what you want to do. And very much AI is a huge piece of this puzzle, and you sit where you are, and you send your senses across and do whatever you want to do. Now, there are some senses they can even smell now a bit. It's not perfect, but one of the ones you want very quickly is touch, haptic hand, et cetera. So very clever in how they did all of this. But the point being that when they decided to do it, they wanted to think about the ethical dilemmas from the very start associated with what they were doing. And they began to work...they worked very hard at talking to a whole series of stakeholders everywhere about what they needed to think about, et cetera.
This has evolved over time. As I said, it was about seven or eight years ago. And they have their giving avatars. It turns out government, the government is supporting it because with the elderly population, people can't get out. So you can be in your home and do what you need to do through your avatar robot, et cetera. So it's been interesting to think about these. They actually also needed to... Well, the CEO of ANA, I mean, the regulated... When they went to him and said, "We think you need to go into the teleportation business." He also said, "You mean like Star Trek?" And they said, "Yes, but it's not going to be the mass." So the reason why they actually invested, but now they have a lot of other investors, and AI has made this all possible in a lot of ways. There are lots of different capabilities you need to actually do this. But one of the things that's very interesting about it is the way they sold it initially was, do you realize that only six percent of the world's population flies in a given year?
Chris Linnane: Okay. I didn't know. I thought it'd be higher than that.
Linda Hill: No, I know. It's just six percent.
Chris Linnane: Wow.
Linda Hill: And it's probably not going to get all that much higher. So how is ANA going to grow? You need to get into the mobility business, not the airline business, and you need to get into teleportation without the mass.
Chris Linnane: Oh my gosh.
Linda Hill: Now, the board, this has all been quite a process, but they have one of their avatar robots. I guess I don't guess it'll be coming down soon, up in the International Space Station. And the reason why construction companies help them with that is because you can use these avatar robots to go into places where you don't want humans to go, but they're controlled by humans from another location. You can shop with them. You can do all kinds of things with these avatar robots. They're not perfect by any means. This haptic hand, they've made a lot of progress on, so you can touch from being in your home and touch something, and there's a tourism business being created where you can fish for tuna from your home, et cetera.
Chris Linnane: Wow.
Linda Hill: So they have all... Yeah, no. And so the technology, and truly these emerging technologies and a number of different subfields, AI being a major player in all of this, who knew that we'd be doing this at this point in time. So I think that the imagination and the lack of constraint that one often has in your imagination as a...anybody can have it. It's not really age-related. That's one of the things that people say, but they've just...in Japan, they are very comfortable with the robots. If you think about some of their pop culture, et cetera, and the kinds of possibilities that they see, tremendous. And I think that when you engage a lot of the work that needs to happen for us to really get the positives out of gen AI and agentic AI, I think are going to be multi-generational, cross-generational collaborations. And they don't happen as much as they should in most companies.
Chris Linnane: Yeah. I never actually thought about the different regions or different nationalities and different locations being more willing to experiment with AI and or robotic things. That puts a little twist on what I was thinking.
Linda Hill: Yeah. No, what I do know is you can look at the capabilities of a country, right? But one of the things that came up is that a lot of people wanted to invest in their business, and many people said, "You've got to get out of Japan." Japan is, at the moment, not really known for innovation in the same way. It's a place unlike Silicon Valley or something.
And so all of these investors were like, "You need to leave if you really want to succeed." They did an analysis of what was in Japan, as well as the advantages of being connected to ANA when you're a startup. They're spun off now, but ANA has invested in them. They said, "No, we need to stay right here in Japan." So the other thing that you need for GenAI to take off and for these bridgers to be able to do what they need to do, it's not just emotional intelligence of how to work with other people. It's what is referred to as contextual intelligence, being able to read what really matters in a given context. So when they did their analysis of which country should they be doing this in, the US wasn't one of them right off the bat, in part because we're not as mobile. We now are, but as you well know, a number in Asia, they sort of jumped to mobile solutions when we're still kind of computer-based.
...jumped to mobile solutions when we're still kind of computer-based compared to other parts of the world. So there were lots of different factors, but the cultural ones about the comfort with robots, they're very comfortable with robots in Japan. And they're also very capable because of the auto industry, et cetera. And these are general-purpose robots, just like general-purpose AI or gen AI. So they really concluded, stay rooted right where we are to get this off the ground and thrive for now. And obviously, they want to become a global entity.
Chris Linnane: So what does AI make uncomfortable for leaders, even very capable leaders?
Linda Hill: There are a number of things, but one thing is that this whole question about whether these new capabilities lead us to have to get rid of lots of people? Now, I don't think there are any senior people who like to do layoffs.
Chris Linnane: I hope not.
Linda Hill: No, it looks like it might to some people, but that is terrible. So this whole question about what will my workforce actually look like? First off, what will the work look like? And what will my workforce look like? These are unknowns to many people. And you and I can read any number of articles that will tell us it's going to look like this or that or whatever. What's the reality?
And when we first begin to ask C-suite executives, "What do you think? Is it going to be a substitute or is it going to be complementary?" Eighty percent of them said they thought it would be complimentary. Within a year, 53 percent said they thought it would be complementary. And indeed, more of them were beginning to think it was going to be a substitute.
Now, part of that is because, and I'm going to answer this from conversations that I haven't completely analyzed, part of that was because the way people, a lot of people are thinking about how are you going to get value from gen AI is by automating lots of work. And then when you automate, you can remove people. But as they begin to understand what it really takes to use these tools effectively, it's not so clear that you remove whole people if you ... A person. You may remove functions, some of the functions that the person did. So a lot of the solutions, and I don't know that I'm saying this in quite the right way, a lot of the solutions are going to be horizontal.
Chris Linnane: Okay.
Linda Hill: So if you have 100 professors, with gen AI, you can figure out which tasks they all have to do that really they shouldn't be spending a lot of time on. So we need to take those tasks, and we can automate those, but you still need the professor for a range of reasons. So it's not so much that you're going to necessarily get rid of a whole lot of people, but those people should be able to work on that work that humans are uniquely capable of doing.
And we don't know what that is yet because we don't even know what this technology is ... The so-called jagged edge, as they tell me over at MIT. So it's still evolving as well. So, trying to figure out how do we remove...what can we take away? But I think it's more, how can I make those who I have already be able to do their work more effectively?
Now, that 80 to 50 percent, that shift was originally, I think a lot of leaders, and it may come back to this eventually, just thought they're going to be able to chop off people. So I was talking to a leader about one organization in which they got rid of about two-thirds of their customer service people because they thought they could use this tool, and they wouldn't...and that's a real cost to get rid of those people. They had to hire almost all of them back because it turns out that the tool they built, it could talk, it could be very sensible when the human was in the first interaction with that one customer. And then the second interaction, it was okay, but it was getting a little...by the third interaction, it was totally hallucinating. So all those people they got rid of, they've had to hire most of them back.
Chris Linnane: Wow. Small win for the humans. Temporary, maybe, but small.
Linda Hill: Yeah, and it may be temporary, right? And there is work that humans shouldn't have to do. And frankly, if you can automate it, automate it. But I think that there's a lot of experimenting that's going to be done before we get there. I do think you can't wait to get there and then decide as a society, we'd better do something about it. I think already, this is not a political conversation, but inequality is not good for any society.
One of the companies we were studying is a company that is an auto company. Actually, Renault. So the person who was the CEO of Renault, Luca de Meo, he wrote a letter to the EU citizens about what it would take if the EU were going to remain competitive in the auto industry. And what he argues is that it's going to take an ecosystem approach. And he makes that argument because he said, "Even if Renault was the...we're the best-run auto company in the world, we made these incredible electric vehicles. And you know what? If we don't have charging stations, it doesn't matter. If we don't have the right rules in place about the compliance regulation for these electric vehicles, electric vehicles is not going to work. If we don't work with NGOs to upskill all the people who need to be upskilled so that we can make an electric vehicle, as opposed to what we were making, if we're not all working together in a more collective fashion, this industry, over time, will die. And the auto industry in most countries is really a fundamental industry for driving economic development."
So, a beautiful letter to the EU describing to the citizens why we all need to understand what's coming. And if we want to meet our goals of sustainability or affordable electric cars, whatever, then we need a different kind of workforce with different capabilities.
So I think that to assume, maybe we're better off having people understand that it could substitute 50 percent, 50 percent might be substitutive work that happens, and we're going to have all these unemployed people. Well, who wants to live in a world where you have a very small portion of your population having to support all of these other people who have no work? So I do think that maybe it's better that they admit that.
Now, one of the leaders who we are studying and is making real progress with AI and gen AI and getting people to adopt it, and I can describe how and why, but one of the things he did, he said, "You know what? Let's don't call it artificial intelligence. Let's call it augmented intelligence." And augmented, okay, it's not artificial. And it's made by people as well. He said, "It's human, and it has all its own different sets of foibles that you have with all humans." But augmented intelligence, even though he knows that there's a lot that is going to be automated, and he wanted them to think about what aspects of your work do you wish you didn't have to do? So working with, this is stuff you don't want to have to do. Okay, should we start there with some of that so that everybody gets to enjoy their work better every day?
So these approaches, and it's not to be sneaky, he's a man of much integrity, it was more like how do I get people to engage to help us come up with solutions that are really going to make a difference? Because there are lots of complaints now that no value is being created by these tools. But before he did that, he decided he had to take a number of courses and understand the technology himself. And he took one course; he had all kinds of exams. He's like, "Oh, I have to go do my homework. I got to take this exam. I got a paper to write." And he's an SVP at a very large company. And he said, "I have to understand it myself more."
But then he made it mandatory that if you were at certain levels or roles in the organization, you had mandatory AI training once a week. And it just showed up on your calendar. I think he got to play professor a bit. He did some teaching of that and had other people help, sometimes different generations, et cetera.
But then everybody had to come up with a use case, a real use case in whatever their function, wherever their level. And people started using it, and you had to look at these things, and they voted on them, et cetera. So the whole organization became engaged in, what is the possibility here? And many of them said, "Oh, I don't really like statistics. I don't like all of this," but got them all very engaged. But then what happened is they loved it, and they wanted to implement all of it.
Chris Linnane: Wow.
Linda Hill: But what he had forgotten to teach them was how expensive it was.
Chris Linnane: Oh, I'm sure that was a surprise.
Linda Hill: So they had another couple of sessions, mandatory sessions on the cost, independent of the electricity, and all the things that… He said, "I forgot to educate them about what it costs." And then they're all like, "Oh, no, we're ready. We have our use cases." But then again, you have to be able to prioritize and figure out which ones should you work on. Again, you need an organization that knows how to make those decisions.
Chris Linnane: So when he put this out to the organization, did he call it augmented intelligence?
Linda Hill: Yes, he did.
Chris Linnane: Smart. It's accurate. I'm more likely to drink juice that has augmented flavors as opposed to artificial flavors.
Linda Hill: Yeah.
Chris Linnane: So I can see why that word feels more like, "Oh, you're helping me do the thing I already do." I wonder if that simple branding situation of artificial intelligence has just rubbed so many people the wrong way so early that it's kind of hard to get that toothpaste back in the tube for most people.
Linda Hill: I think so. The other one that relates to it, and I mentioned it a little earlier, was this whole idea of...well, I don't think I said a whole lot about it, but we found another thing very similar to that is when you say data-driven, it turns out that data-driven, which is the word that most organizations use, first off, it's inaccurate, but it also makes people crazy. So when you say, "My decisions are supposed to be data-driven," it implies that my experience and expertise and my judgment are not as valuable as the data. And since who came up with this data, those weird-looking people over there in the IT department or whatever. It really led to a tremendous amount of resistance, particularly in professional service firms.
And hospitals, for instance, doctors like, "Well, wait a minute, I've been taking care of whatever." Now, I know that I saw something on the radio this morning that said, "Soon we'll have no doctors because AI is going to be able to make better decisions." But AI is not going to make you feel okay, right? When you're sick, there's a whole emotional component to this that I don't think even some robots aren't going to be able to help us for a while, particularly when you're really sick, but they should use the tools.
But what I was going to say is that this idea that it's data-driven, we found in our own research, made people crazy. So one of the leaders we were studying, more than one actually, began to use the word data-informed instead of data-driven. Now, I want to be informed by data. I'm a logical person. So okay, data-informed. Now, what data will inform me? That doesn't imply, and it implies judgment. And frankly, now that we are using gen AI, you really want people to understand it's simply data-informed because you do want them to ask questions about what data was used, who did this, where does it come? Does this result look like it even makes sense or it looks like it makes sense, but let me dig a little deeper. So you actually do want people to use their judgment, expertise, and experience, that tacit knowledge to figure out what they're going to do with the solution that came out of, again, the algorithms.
So I think we have really advocated for data-informed as opposed to data-driven. We need to abandon that word. Now, I think the genie may be out of the box, even with that, for all I know, because everything says, "Are you data-driven?" But it's actually inaccurate, and it now truly matters that people understand the data does not give you the answer.
Chris Linnane: Is it too late for me to start a PR agency where I rebrand AI and I take over the way that people accept this? It's probably too late. I won't try that.
Linda Hill: Maybe not, have faith.
Linda Hill: If teleportation can happen, okay?
Chris Linnane: Yeah, if teleportation can happen, I can start-
Linda Hill: Yeah, you can do-
Chris Linnane: ... a PR company for AI, the thing. Interesting. It's the way I haven't looked at it before.
Linda Hill: Yeah, and it's funny because I think one of the...again, going back to leaders, I have so much to learn about human-AI interaction or collaboration. But one of the organizations we studied was Cleveland Clinic Abu Dhabi, and it is an extremely...the digital transformation was done really in a beautiful fashion. And really, the groups that were involved in it were, they had the digital teams, they also had a business intelligence team, and then they had an organizational effectiveness team, and they all worked together. And they started by saying, "We want everything to be very user-friendly and as much as possible, whatever we create, the individual should be able to use it themselves." Self-service.
So because they said, "That is our goal," creating a digitally dexterous workforce as that one that's going to be willing and able to embrace and use these tools to make a difference, because that was their goal, they then, working backwards, said that, "We need to involve the users," what they call them caregivers in the hospital, anybody who works in the hospital is a caregiver, "We need to involve them upfront in the design of what we're doing and also the deployment."
So one of the things that was fascinating to me in this situation is, as with all organizations, when they're doing these things, they have dashboards. And one thing that began to happen is that the caregivers were sort of calling the digital team or the business intelligence team and saying, "Why did you design the dashboard this way? Well, I'd rather have these data than those data."
Now, at first, of course, it's expensive, these calls, et cetera, but when I was talking to them, I said, "That is a sign that people are deeply engaged. When they actually call you, they bother to pick up the phone or whatever, send a message saying, 'I would prefer if you had this data, because this data would really help me make some decisions I need to make.'" So I was like, "Well, wait a minute. Who are these people calling and telling us how to design our dashboard?"
Now, one of the dilemmas they came up with and they resolved it over time is that you do want standardization of some of this so that you can have the collaboration across, et cetera. But that is how engaged people were because of how it was designed. They own it. They feel like they own all of it. As I said, they run A/B experiments. They all know what A/B experiments are and how you can design them so that you...and what allowed them to do it was visualization tools, which are really important. When people can visualize, you get much more adoption of these various technologies.
So they actively were engaged in pushing and pulling on, why did you deploy it this way? Or did you ever think about doing it that way? So once you have the population you're trying to transform become champions and deeply engaged and not arm's length about that transformation, it's a whole new ballgame. So I think that when you were asking in one of your earlier questions was, well, what do we need to learn differently about...we're asking. And one of your earlier questions was, "What do we need to learn differently about how to do digital transformation now that we have AI?" If anything, we can learn from those leaders who've done it the right way. That we really need to figure out how to do it in a way where people feel quite empowered about the design and delivery, deployment of it. Then, we have typically done it.
Chris Linnane: So, from almost like a backed-up view, where do AI initiatives often just stall out?
Linda Hill: I don't know that I have enough data to answer that question, really, but what I will say in answer to that question is you do want to focus on making... Whatever you do and however you do it isn't worth the investment unless people are going to be willing to use, and able to use, whatever it is you're working on. So if you take that very seriously and don't just focus on what, but the how, you're going to get there. I think you're much more likely to make progress. Because often, with all the new innovations in general of any sort, there's so much focus on the what, and not enough focus on the how and the why. And that's why people are...
We're human. We want to understand. We want to...and we don't like to take too many risks, a lot of us. And we don't want to look stupid. And all of those things, we know. And the leaders who realize that they're dealing with people with emotions, and that AI, in particular, all kinds of emotions. That people have some, they're exhilarated. Some, no, not so much. You really need to pay attention to that. So in our article, we've written about why are some companies having more success in adoption. We see that one thing that is true is that the senior management, the very senior management, is engaged and they're bridgers. They're actually interacting with people throughout the organization. They know everyone's not going to come along, but they're really talking to people about how they feel about it. They're actually talking about emotions, and it's not just an intellectual conversation. And so they take that seriously, either in informal settings or not so informal settings, and are helping people understand why it might be worth the investment.
So as an example, one of the things...as I mentioned to you before, the leader was making mandatory AI training. Before he did all of that, he talked a lot. He developed a story, a narrative that he deeply believes. So, it's not about why they need to do this. And he's in the pharmaceutical industry, and he said, "If we want to bring hope to patients as fast as we can, then these are tools that will help us do that. And we're all about helping people live healthier lives and better lives. So if somebody has cancer, and we can use these tools and they get better a year faster than they would have because of what we've been able to develop, is that meaningful?"
So they actually do talk about the why, and they link the technology to a why that is deeply human: purpose. So when we wrote our first book on innovation, one of the things that popped out in 2014 was, "Purpose, people, it has to be worth it." And you have to help me understand why I should be doing all of this because I'm taking risks, and I might even, over time, be sacrificing my job. And I was not surprised in some ways. When we were writing about... Believe me, I'm not the first about anything. Everything's already been done. But since we wrote Collective Genius, and it came out in 2014, and we said very upfront, "There has to be a clear purpose." It's not so surprising to me that in the last, I don't know, three, four years, five, there were all these books about purpose. Because we're asking people to take risks and to be involved with technologies that might make them redundant or certainly will make them feel very at risk for some period of time.
So I think those leaders, that's another thing they do. They actually explain, "This is an enabler for us to do what we really think is most important to do." Similarly, in financial inclusion, an executive said, "We're going to be about financial inclusion. More people will be able to be involved in the formal financial system if we can become more of a fintech, because it's more affordable if we get that in place." That's meaningful to people. Okay. I really have always felt pretty guilty that we haven't been able to support those small businesses in our neighborhood. And if we could get the price point to that because we use some of these technologies, that's why I got into what I'm doing.
So, I think being able to understand these leaders do engage people at the level of the emotion, the emotional challenge of the work. They don't act like there isn't an emotional challenge because there is. They might say to themselves, "I don't understand much about this." They show some vulnerability themselves about whatever it is, and I'm trying to learn. And what's interesting to see is when the younger, the digital natives, reported to us about how they perceived the C-suite, when they saw the C-suite actually come to some of the hackathons, be involved in some of the rankings, et cetera. Their sense of the company, their sense of trusting management, and that management's trying to do the right thing and understanding the... Getting to know that they're people, too. As one CEO said, "It really does build a different level of trust." And I think when you're asking people to do a lot that is risky...I hope he'll forgive me for this. He's really happy in his job he has now. He was out of college less than six months when he got laid off.
Chris Linnane: Oh, no.
Linda Hill: Because the tourism was down in the US, as you know. He had just gotten a promotion, and then he got laid off. "Mom, I'm supposed to go to the Unemployment Office, and I'm not even out of college for six years." He's now in a wonderful opportunity. Just a few months later, I said, "Let's hope this is the only time in your life you're laid off." But I think that we need to deal with the fact that people are anxious, people have legitimate concerns about privacy, about what kind of world is this going to create. And you can't ignore those, and you shouldn't. And when you engage them, it doesn't make you look weak. It makes you look like you're thinking about what is really on their minds.
Chris Linnane: Can we try an exercise?
Linda Hill: Mm-hmm.
Chris Linnane: All right. I'm going to read you some questions that leaders often ask about AI. And these all sound reasonable, these questions, but I think, I think, these might be maybe the wrong questions. Potentially askew, maybe wrong. I don't know. So, I'm going to read the questions. And you don't have to go into much detail, but when you hear it, if you think it's appropriate to reframe it a different way, I'd love to hear what the question maybe should be instead of this question.
Linda Hill: Okay. Not answer it, just what it should be.
Chris Linnane: Yes. What you think is maybe a more useful question to ask, because these questions may be the wrong way to think.
So leaders are asking, "How fast should we roll out AI?"
Linda Hill: That's not framed incorrectly.
Chris Linnane: That still works?
Linda Hill: That still works. How fast? Yep.
Chris Linnane: All right. Yep, how fast? Next one, "How do we get our people to trust AI?"
Linda Hill: The question is more, before people trust AI, what needs to be the case?
Chris Linnane: What do you mean by that?
Linda Hill: It depends on...there's lots that have to be in place before I should trust it. Where did the data come from? Are there biases built into the data, either inadvertently? Because I don't think you normally build them intentionally. So, what is the maturity of the particular tool we're using? What does it take to actually use it properly? Have you helped me understand how to actually interpret what the output is going to be? I think there is a whole series of questions to be asked.
Chris Linnane: And earn the trust.
Linda Hill: And earn the trust, yeah.
Chris Linnane: Okay, next one. This is interesting because, yeah, I like to hear what you think about this one. Just asking the blanket question, "Which AI tools should we standardize on?" Is that potentially a problematic question?
Linda Hill: Probably.
Chris Linnane: And why do you think?
Linda Hill: It should be standardized on because it really goes to how should the work be done and where should the decision-making rights be. So, you don't want to just...AI cannot solve the more fundamental question of who should be making the decision? So once you understand who should be making the decision, then you can maybe use the tool to help you with that. Because one of the things we find, we have an instrument to assess how, whether or not you have the right kind of decision-making process and norms to allow you to make decisions about innovation, either solution or deployment. And most organizations are kind of broken there. So if you just take what you're already doing and just use AI to help you with it, it's going to be ineffective.
Chris Linnane: Okay. Last question. Is this a good question, or is this maybe the wrong way to ask this question to your organization as a leader? "Do we have the right AI talent?"
Linda Hill: Do we have the right AI talent? I guess it could be, but I guess everyone is going to be using it to some extent. So I guess, I would say more, are we...a couple of things. Going back to it, it's not just individual talent; it's how the people work together. Because what needs to happen is they need to be able to collaborate, experiment, and learn together. And it makes me think about when Ed Catmull from Pixar came to speak to the students. We wrote...our first chapter in our last book was about Ed. Ed came, and the students said, "The reason why Pixar is so successful is that you have more talent than most people." Or most organizations.
And he said, "We do have a lot of talented people, but I don't think that's the reason why." He said, "I think it's because we understand that filmmaking is a team sport. And we know lots of teams, lots of talent, and they don't do very well." So, he described what he meant by filmmaking being a team sport. And after he did that, someone said, "You can't lead that way, because if you led that way, you'd have too many cooks in the kitchen." In other words, too much talent almost.
He said, "You can never have too much talent. Never." So the issue is you, as a leader, need to figure out how to get those top chefs to be willing to cook a delicious meal that someone's willing to pay for, and that's your job. But as long as you think you can have too many cooks in the kitchen, you will never have a company that can sustain innovation over time.
Chris Linnane: That's awesome.
Linda Hill: I was torn about the fast one, but you know speed matters. And we hear from everyone that you want to have a bit of that first-mover advantage, usually in the tech world, or when these tools are very important. So speed does matter, so to ask how fast. But to understand, of course, the risks associated with moving fast. You need to be clear about those and try to mitigate those risks.
Chris Linnane: I think we can wrap it up with this question. So, given everything we've talked about, what should leaders focus on building first if they want to be ready for AI and not just deploy it?
Linda Hill: To really leverage AI to make a difference, create value, whatever kind of value you're trying to create, you actually are going to have to get people to be able to work differently. So you need to step back and ask yourself, how do people need to work if, in fact, we're going to be able to leverage these tools? And if you can't answer that question, then you have a problem. And then why people are working the way they do, and how you're going to have to transform the way they're going to work if they're really going to use these tools? That whole process is going to take time. Culture, even in a medium-sized organization, doesn't shift quickly.
So I think that you need to really step back and understand, how are we going to have to work to use these tools? And do we work that way? No. Okay. Why don't we work that way? So on the one hand, you can go back and analyze because of our performance management system, our career management, our structures, whatever it might be. But okay, if you don't have time to do everything and you can't do everything at once, but do that diagnosis before you act to understand why there would be, in some ways, legitimate resistance.
Because often, we think that when people resist, they're being unreasonable or they just won't change. No, they often have very reasonable reasons for resisting because there are costs associated with that change, and people do not change until they see that the benefits outweigh the costs. And it's about that person's perception, not your perception, of the benefits and the costs.
I think it means you need to do some analysis and make sure you understand that the tool is not a panacea. It's not a magic bullet. If the work is still designed in a way that people aren't going to collaborate, then they're not going to ever leverage that tool, for instance. So, what do you need to do to get them to collaborate more?