How can you, CEO or other senior leader, harness generative AI as a co-thinker and strategist?
In this episode we'll share cutting edge insight about how you can make that happen.
Elisa Farri is vice president, co-lead of Capgemini Invent’s Management Lab. She is listed in the Thinkers50 Radar Class of 2023. She is a frequent contributor to leading management magazines, including the Harvard Business Review, Forbes and Rotman Management. She has recently published the "Management GPT" white paper in collaboration with Thinkers50.
In this conversation, you’ll learn:
- Three impactful ways leaders can start using generative AI for, as a thought partner
- Why you need to move from knowing about generative AI, to actively harnessing it
- How to use AI as a co-thinker, not as a producer
- Where you should start
Resources/sources mentioned:
Join Rivendell (https://xquadrant.com/rivendell/), our exponential programme for elite CEOs dedicated to transforming themselves, their businesses, and the world.
Ready for a big leap forward in your own leadership? Elevate your leadership forever with Richard's book "Making time for Strategy: How to be less busy and more successful". Buy your book here: https://xquadrant.com/time/
Subscribe to The Impact Multiplier CEO Podcast
Watch
More of a video person? No problem.
You can watch this episode and discover more videos on strategy, leadership and purpose over on the Xquadrant YouTube channel.
Elisa Farri
So the way in which I see the evolution of generative AI in a way that is embedded, it’s not about the technology. It’s about we as humans doubling down. So we need to double down on some of our human core competencies and skills, like the art of asking questions, strategic thinking, judgment, problem solving, and using the The machine in a way that can truly help us exercise our capabilities.
Richard Medcalf
Welcome to the Impact Multiplier CEO podcast. I’m Richard Metcalfe, founder of Xquadrant, and my mission is to help the world’s top CEOs and entrepreneurs shift from incremental to exponential progress and create a huge positive impact on our world. Now that requires you to reinvent yourself and transform your business. So if you’re ready to play a bigger game than ever before, I invite you to join us and become an Impact Multiplier CEO. How can you, as CEO or other senior leader, harness generative AI as a co thinker and a strategist, and actually pull away from other leaders in this area? Well, you need cutting edge insight if you wanna do The. And today, I’m proud to bring it to you. Alisa Farri is vice president and co lead of Capgemini’s Invent Management Lab. She’s listed in this Thinkers 50 radar class of 2023.
She’s a frequent contributor to leading management magazines. And she’s published a few key white papers that have been seminal in this whole area of how do business leaders use generative AI for real business outcomes. So it’s really easy to use generative AI to write a blog post or write an email for you. But What we wanna look at here is how to use the technology to be a thought partner in leadership and strategy. So this is a fun conversation. We look at 3 key use cases that Alisa has identified as being the key ones that he did should start with. We look at why you need to spend more time using it and experimenting than you are at the moment. And we talk about, practically, how should you start doing The? So this is a great, interesting conversation on a really important, new trend that we’re also aware of right now.
Enjoy this conversation with Alisa Fari. Hi, Alisa, and welcome to the show.
Elisa Farri
Thank you, Richard. My pleasure.
Richard Medcalf
So I’m really fascinated about this conversation. I know you’ve been researching for several months, perhaps even years at this point, but perhaps not too many years, because it hasn’t been out that long. But the impact of generative AI on leadership and management, I know you’ve already written, a number of pieces for the Harvard Business Review and a recent piece on management GPT, in collaboration with Thinkers 50. So I’d love to kind of dive into that. So but, Pat, first of all, set the scene. Like, how why did you decide to jump into this topic of generative AI? Obviously, it’s very hot, but what excited you about it?
Elisa Farri
Thank you. Thank you, Richard. So my journey of experimentation with generative AI and in particular The the JGPT started almost 1 year ago. Maybe also for your audience, a couple of words about who I am. So I’m a coleading Capgemini Invent Management Lab. And so we are specialized in some management practices and theories, and our work is to bridge them into practice. So to take them to the real world. So we we do this work at the intersection of the academic, the The, thinker in the world and the business the business world.
And so you might ask yourself, so it is a what’s the link, you know, between what you do as managing a lab and generative AI? So it’s Impact when you’re good together with my team, we sat together and we started debating, you know, whether generative AI was not topic for us to cover or not. And I must say that the beginning, I was very, very skeptical. So I need to thank my sparring partners at the management lab, and they are Apollo Savini and Gabriela Rosani because they pushed me a bit out of my comfort zone, asking me, you know, please, Lisa, get your hands dirty, you know, practice a bit with CIGPT. And then, you know, let’s reconvene. Because, you know, we have a feeling that this might be relevant to us as well. And so I must admit that, you know, the very moment I started, let’s say, using, ChogpT, I realized the big potential, you know, of this technology, the first technology that can be associated to other similar, let’s say, consumer technologies. So easy to use, an interface that allows you to have a human like conversation with The machine. Nothing like this was accessible before.
And so in our experimentation journey at the management lab, we said, why don’t we start using the machine in our own domain of expertise? Expertise? Okay? So because we wanted, you know, to, challenge the machine, and so we said, okay. Let’s start from what we know. And, and so the first big experiment that we run was in May last year. We asked the Cajcipiti if it was knowledgeable about 10 of the most popular management theories and practices and then enter The into, let’s say, sort of a basic conversation, basic dialogue, which was just a sequence of questions that we asked the machine about ways in which generative AI could announce the application of these theories or even disrupt these theories. The answers that we got were fascinating. So we realized and we said, wow. The machine knows The theory is very, very well. So it’s very proficient.
He’s a black belt in this, in The series. And so the results were so fascinating that we published them in a ebook together with Harvard Business Review, Italy, both in Italian and, in English. And and the book got viral very, very quickly. And the reason was very simple. It was the first of its kind The targeting, management topics and generative AI. During the summer, then, you know, to also to walk you through our journey, what happened was that, you know, we were approached by many experts, also curious, friends, and partners, you know, asking about our experiment and also challenging us a bit, you know, asking, what’s next? So how can we truly unlock generative AI potential in helping managers, you know, apply, these theories in their daily work? And so we sat together again as a team at the management lab, and we started brainstorming about how to go from a text written in a book to, certain prototypes. And so that’s how, the management GPT idea came, came to life.
Richard Medcalf
Great. So, so let’s take management GPT. I know one of the themes, for example, is around leadership, responsible leadership. So, yeah, if if I’m a senior leader in a business, like, I’ve my my subscription to whichever service, you know, I use, But why would I even wanna go and start having a conversation with The chatbot around my own leadership? And and The how would you recommend I start doing that?
Elisa Farri
Well, this is a great question. And to be honest, this is the typical question that I got over the last 6 months when I had conversations with CEOs, The CXOs, and managers at all level. Before I jump into The specific case of responsible leadership, I would like to share a couple of numbers with the audience. The first one is that it’s a no brainer that CEOs and top executives globally see a big benefit in this new technology. Okay? So more than 96%, according to our Capgemini Research Institute and latest research, say that this is a hot topic that they discuss at boardroom level. Okay? But there’s a but. The point is that only few of these executives are using this technology on a daily regular basis. And so in the words of professor Karim Locani of Harvard Business School, we are seeing a practice doing gap, a gulf that is growing between people who are using the technology, getting familiar with the technology, seeing the benefit, the risk, and others who are not even aware of how the interface looks like, and so debating about a technology without having direct knowledge.
And so there’s a a a urgency now to close this gap and help leaders in every organization being more familiar with the technology because without The hands on knowledge of the technology, the risk is that they might take decisions that are not well informed. And so starting from this as a big issue, we identify the 3 use cases inside of this new publication called management deputy. One of them is responsible leadership. Why responsible leadership? Because together with Edward Brooks and Corey Preston, who are part of the Oxford character project, we sat together and we discussed about one of the challenges that executives are facing, also in the pre generative, AI era, about having self reflections about their character, and so how they behave, The habits, positive, negative, the actions they take in order to be true role models of good responsible leadership. And so in our journey of experimentation, we started using generative AI models as sparring partners for executives, helping them ask the right questions, walk them through a structured dialogue in which, I want to clarify, is always the human, is always the manager answering. It’s always the manager making decisions. The machine is like your smartest aspiring partner in giving you food for thought, example of proven techniques, highlighting some overlooked aspects and dimensions, suggesting habits that you may want to start practicing day by day in order to embody responsible leadership. So the point of using generative AI is not absolutely to replace human coaches or people.
It’s about giving executives a neutral interface, which can be a black belt from a methodological standpoint and guide them through an in-depth powerful self reflection.
Richard Medcalf
Where would somebody start? Like, in your paper, do you just list out like, do you type in these prompts 1 by 1, or do is it more of a flowchart where you’re gonna or is it more just, like, start a conversation. Here’s a good opener. How do you kind of structure these conversations?
Elisa Farri
So another another great question. So, yes, you need to structure the conversation. You need to architect first the type of dialogue that you would like to have with the machine. So together with Edgar Brooks and Corey Crossan, we sat together, and we defined the flow of the conversation based on the actual character project research, their own past experience, and academic, publications, and then gave this as a sort of instructions to the machine. And then, within Capgemini Inventor, we had an Indian team developing a customized bot powered by GPT. So the users now who are testing the bot do not have to inject or copy and paste the prompts. It’s like having a, predesigned bot asking you the questions and engaging you, in the conversation. But this is an important message for people and for executives.
So you need to build and exercise this muscle of thinking, designing, and engaging in effective conversations with the machine. Because if we go back to the practice doing, the knowing doing gap of professor Karen Lakhani, there’s a deep root cause behind this. Most of the narrative is about using the machine, in a consumption mode. Okay? So you use the machine, you ask a question, you wait for the answer, you ask for a small task, you write me the email, and that’s it. So pure execution. In some cases, even full automation. But the machine can be used hosting another interaction mode, which is what we call the the co thinker. And so if you tell the machine the type of questions that you would like the machine to ask you, the machine is very, very good, you know, at following these steps by steps.
But, again, this goes back to our strategic thinking, to our judgment, to our ability to strategize the type of conversation that we would like to have with the machine and then run up our sleeves and engage into the conversation because this is the other point. So we’re seeing, a lot of people who start, using The models, both company customized models or external third party models. And then, you know, they get a bit either lazy or they simply drop out. Why? The answer is simple. It’s because they they understand that it takes a lot of time and effort, you know, to give the context, the information, the details to the machine because the machine does not know who you are, your context, The organization. And so you need to guide the machine in a way that you can truly help the machine raise the bar of the performance. And so this goes back to our role in how we want to interact in a responsible and safe way with the machine.
Richard Medcalf
I hope you’re enjoying this conversation. This is just a quick interlude to introduce you to 2 transformative programs that we run. The first is Rivendell, my exclusive group of top CEOs who are committed to transforming themselves, their businesses, and the world. It’s an incredible peer group and a deep coaching experience that will push you to new heights no matter how successful you’ve already been. The second is Impact Accelerator, a coaching program for executives who are ready to make a big leap forward in their own leadership. It’s regularly described as life changing and no other program provides such personal strategic clarity, a measurable shift in stakeholder perceptions, and a world class leadership development environment. Find out about both of these programs at xquadrant.com/ services. Now back to the conversation.
Yeah. That’s great. So if somebody’s thinking, oh, this sounds fun, how to experiment this, do they need to build a bot to do this? Do they need you know, how how can somebody practically start to get value from what they already have?
Elisa Farri
Okay. So so, obviously, we built a bot because we wanted to run an experiment at scale. My suggestion for individual executives and individual chief executive officers in particular would not be to build a bot. And, let’s be honest. I don’t think that in the future, we will have, like, dozens of bots on our desktop that we can activate, you know, for each specific managerial task. So the way in which I see the evolution of generative AI in a way that is embedded into, managerial task is more as an interface that you have, a conversational interface, human like, type of conversation interface that allows you to interact with the machine. And so, again, it’s not about building a bot. It’s not about the technology.
It’s about we as humans doubling down. So we need to double down on some of our human core competencies and skills, like the art of asking questions, strategic thinking, judgment, problem solving, and using the machine in a way that can truly help us exercise our capabilities. And so it’s, personally, I feel that this is the dawn of a new capability that all managers will need to master in the future. And the risk that they see is that managers who are not investing now in building this capability will risk, you know, a career misstep in the future. And so we risk some inequality, some goals between managers who are using the technology in in an effective way and those who are not. And and the positive thing about this technology is that it’s not always, is not only easy to use, but it’s also accessible, notwithstanding your background, your powers, your geographical location. And so it’s it’s very accessible and inclusive. And so I hope that, you know, by raising awareness that the machine can be used in various ways depending on the task will help managers work better in the future.
Richard Medcalf
Mhmm. Yeah. Yeah. I can see that. The could they were they’re all parting of the ways is the skill to build. I’m kinda curious. Yeah. Where might somebody again, where where might somebody start? Because, again, you’re right.
I mean, there could be a whole load of conversations to have with a with an AI. Would you start with, like, a very tactical thing? Like, you know, something that’s happening in your team? You wanna, like, problem solve, or would you go into more of a reflective kind of, conversation around, you know, help me explore my leadership values or whatever? You know, I’m just wondering where you might start because it will take some time to experiment and to craft prompts and to type your own, you know, half formed thoughts in and so forth.
Elisa Farri
Absolutely.
Richard Medcalf
So where where might somebody want to start if they wanna kinda start to go up the learning curve curve as it applies to their own management and leadership?
Elisa Farri
I would suggest to start from a topic that is of interest and those are topic that person masters. So you feel that you are the expert in in the topic. Because especially at the beginning, you know, at the of the learning curve, you want to be 100% sure that you can criticize, that you can judge, or you can even weight the output of the machine gives you. So for instance, I did not start, you know, from, I don’t know, medical or health care topics. You know? I’m not a doctor, so I wouldn’t be in a position to say if the machine is giving me hallucinations or not. And so in my work with, managers at all levels, I always start from their own expertise. So you’re a manager working in the finance unit. You’re working The HR unit.
So let’s have a look at the managerial tasks that you run every day. Let’s separate them. Tasks that can be, let’s say, either semi automated or automated by the machine, and so you will go for the pure assistant, the pure execution model. And so let’s see how the machine can help you be more productive, be fast. Yeah. And you need to separate the words. And then there’s the other part of the world with the list of managerial tasks that are much more complex. So task in which you can’t push a button and then wait like this, and now the machine is giving me the business case.
Forget about it. The machine will never build a business case for you, will never define a strategy for you. But the machine can know very, very well the theory of professor Roger Martin and walk you through the key questions that you should ask yourself if you want to enter a new market. And, and so my suggestion is, once again, start from what you know, start from something that triggers curiosity. It could also be a problem. You know? It could also be something The it’s an issue that you’re facing right now. In other cases, with other managers, what I tried, I asked them to pick a project that, is already closed, CEO something that belongs to the Podcast, also to avoid the The that this is something ongoing, so there could be some psychological barrier. So I said, let’s pick something back 10 years ago.
And now let’s see, you know, how it could change, you know, the way in which, the project was designed and implemented if we start in in with the machine and asking the machine, you know, in terms of feedback and the support throughout the main phases of the project. And this was an eye opener for most of them. And so I remember The project, it was about, let’s say, access to safe water in the salons of Mumbai. And at The point in time, we had to identify the key stakeholders that we wanted to engage into this complex challenge, in this into this complex framing. And, once we injected, we we we share with the machine, the project context, the goals, and The details, the machine performed in an excellent way, giving us a full list of all the stakeholders. And not because the machine is intelligent. The fact is that the machine is trading on such a a huge amount of data that if you can get the machine the proper way, the machine can give you back a lot of, information that can be fit with your goal and The your needs. But then it was again up to the manager judgment to say, okay.
I will prioritize this, and let’s continue the conversation. And so, this is what I call a tango dance. So it’s not a solo. It’s something that you need to play and dance with the machine. Sometimes the machine is on the driving seat because you want the machine to take the lead because it’s a high performer on some subtasks. On others, you want as a human to sear the conversation. And so this should be well orchestrated.
Richard Medcalf
Yeah. Really interesting. I mean, I think The thing I’m taking away is, obviously, one of the advantages of the machine is the access to, like, all the management theory and all the ideas. Again, bringing that point to the, you know, knowledge is free almost. Right? There’s just so much of it, and it’s becoming even closer when you can say, you know, what’s the best management theories for this particular issue I’m dealing with? Walk me through The. Make it some kind of interactive worksheet on steroids. Right? Because it’s literally walking you through the steps. As well as coming up in my mind is actually then the the flip side, which is actually not for the AI world probably, but is actually as humans how we master our own emotional nervous system and response because actually when the when the kind of the intellectual part of management becomes easier to attain, if you like, then it puts the focus on how how we’re actually doing in the moment when colleagues are triggering us, right, when we’re in difficult situations in meetings, right, where things are blowing up. I was just on a a retreat myself, a community that I belong to, and we had a day. And the whole point of the day was to actually put us into highly triggering stressful scenarios. And you start to find out what your default responses are, you know, and they’re not normally you at your best. Right? Because you you’re being triggered in in different ways. And it’s really, really fascinating and very, very, helpful growth. And I’m just kinda thinking that for for managers, The are these two sides to it.
Right? There is The kinda, like, the the strategic, like, okay. How am I dealing with these issues? And then there is the emotional side.
The the embodied experience as a leader and how we show up, which, which I think kind of, again, shows you The this kind of, dance as well between the AI and the human side in being a successful leader.
Elisa Farri
Absolutely. Absolutely. I think that you summarized in a very clear way. And, you know, when I hear or read, you know, that, you know, generative AI or AI in general replace or substitute, you know, manager. In some cases, it’s even managers. So I smile because The personally feel it will not happen, and it should not happen at all. So it’s our duty, our responsibility to make sure that this scenario does not realize in the future. And and for all the reasons that you just shared, so there are there’s the other side that the machine will never be in a position to to replace the The human side The is soft elements, purely emotional intelligence, uniqueness of us is the human.
And plus the machine does not have all the context, the the body language, the the things that we don’t say, but it that we can perceive as humans. And, so, yeah, I think that most of the narrative now is about responsible AI. I think that first of all, we need responsible humans using AI.
Richard Medcalf
Yeah. Beautiful way of saying it. So thank you so much for this conversation, Lisa. It’s been really great to explore this and get into it. I know that the, you know, your white paper, is out. It’s called Management GPT, and I think you covered 3 3 use cases. We’ve got into one of them today around responsible leadership. I know there’s also platform business models, as well as complex problem solving and and how AI can apply to all of those. That’s fascinating reads. I do recommend that for anybody who’s, who wants to go deeper.
Elisa Farri
Thank you. Thank you. And it can be downloaded for free, the publication.
Richard Medcalf
So We’ll we’ll put we’ll put the link right into the show notes so people can find that just by going just go to the show notes for this
Elisa Farri
Thank you.
Richard Medcalf
Episode and link there. Well, it’s it’s been great to talk. Thanks so much. Thank you for pioneering the way in this area, and look forward to kind of seeing, where the thinking evolves over the next couple of years. Thanks again.
Elisa Farri
Thank you, Rachel, for having me. Thank you all.
Richard Medcalf
Well, that’s a wrap. If you received value from this conversation, please do leave us a review on your favorite podcast platform. We The appreciate it. And if you’d like to check out the show notes from this episode, head to x quadrant.com/podcast where you’ll find all the details. Now finally, when you’re in top leadership, who supports and challenges you at a deep level to help you multiply your impact. Discover more about the different ways we can support you at xquadrant.com.
**Note: This transcript is automatically generated.
Please excuse any errors.
Beyond the podcast...
Once you've subscribed to the podcast, why not go deeper and subscribe to the Xquadrant Insider?
This is our complementary email newsletter that focuses on multiplying value and impact at the intersection of leadership, strategy and purpose. Originally designed for our private clients, we've made this available to a wider audience of high-achieving and purpose-driven leaders.