Winning with AI: Charlene Li on Superhuman Leadership and Navigating Transformation’s Challenges (MDE660)

Speaker A: When you think about businesses that are selling through the roof, like Skims or Allbirds, sure you think about a great product, a cool brand and great marketing. But an often overlooked secret is actually the businesses behind the business making, selling and for shoppers, buying. Simple for m millions of businesses. That business is Shopify, it's home of shop pay, the number one checkout in the world. You can use it to boost conversions up to 50%, meaning way less carts going abandoned and way more sales going through. To checkout, upgrade your business and get the same checkout Allbirds uses. Sign up for your $1 per month trial period at, uh, shopify.com income all lowercase go to shopify.com income to upgrade your selling today. Shopify.com income foreign.

Speaker B: M hello and a very warm welcome to all the new listeners of the Minto Dialogue podcast and a welcome back for those who've been loyally listening on today. I am very excited to bring you episode number 661 and it's with Charlene Lee, New York Times best selling author multiple times longtime analyst and entrepreneur, uh, whose work has shaped how companies approach disruptive technologies. Bringing her back onto the podcast, we dive into her latest book, Winning with AI, co written with Dr. Katja Walsh, unpacking the challenges organizations face as they integrate AI and the practical steps for creating real value. The conversation focuses on everything from harnessing AI in everyday business, building trust and transparency, and using AI to foster more empathy and humanity at work. If you want actionable strategies and a fresh perspective on thriving in an AI driven world, this episode is for you.

Speaker A: Charlene.

Speaker B: Charlene. That must be some song Charlene. But Charlene Lee, you are a repeat guest on my show and uh, at this occasion it's great to have you back on. You have a new book but for those who don't know you yet, Charlene, who be thou? Who are you?

Speaker A: So good to be back. Minter. I am a New York Times bestselling author of seven books. Can you believe it? This is my seventh and I have been a long time author and analyst. I uh, was at Forrester Research for about a decade and after that started my own analyst firm that was very disruptive called Altamira Group and sold that in 2015 and I've been bouncing back and forth for many years for corporate jobs and then about three years ago went back to my entrepreneurial roots and became a solopreneur again. So very happy to be having that flexibility and ability to just play where I want and been very focused on this brand New thing called AI.

Speaker B: Yeah, well, the brand new version of it anyway. And you're based, you're based at sort of, let's say, the heart of tech.

Speaker A: Yes, I'm right in the middle of San Francisco and uh, so it is, it fills my, my vision all the time. I go into coffee houses, the farmer's market, and it's what people are talking about wearing physically, um, very visibly oftentimes and it just feels like non stop chatter about it.

Speaker B: I had this vision, uh, when you said farmers market of the farmers actually talking about AI but no, you were referring to the people go buy the good, good vegetables and stuff like that.

Speaker A: Right. You think that that would be one place where you could escape AI but no, that's not the case.

Speaker B: No. And plus, I mean farmers are now in the game as well, right?

Speaker A: Very much so.

Speaker B: Very much so. You co wrote this book with Katya, uh, Dr. Katya Walsh, um, who's got an illustrious background as well. Tell us about the inspiration, the moment where you decided you needed to write this book. Winning with AI.

Speaker C: Right.

Speaker A: It was Ah, 2023 and it was beginning to gather a lot of steam. And the thing that I have historically focused on are these disruptive transformational technologies. And it was becoming very clear that AI this version of again to your point, uh, was a game changer. The fact that you could interact with it through a uh, chat interface made people like me, who are not technical able to use it. Anybody with a browser who could command language could use AI now. And I was in Boston talking to Katja about my desire to focus on this area and to write a book. And she said I should write this book with you. And I'm looking at her going, of course you should write this book with me. I mean she was one of the first chief AI officers in the world. And uh, she was very active in Vodafone, uh, uh, as her Chief Data officer was at Levi Strauss, as her chief AI officer and data officer was at Harvard Business School and now was at Apollo. Uh, and the fact that she had that practical hands on experience, uh, building, implementing and transforming organizations with AI and my experience in writing a book, and she had always wanted to write a book. So we just looked at each other and on top of that we're very dear friends. And so the idea of working on a book with her, which is a major endeavor, excited us, um, together that we wanted to do this together as

Speaker B: a project and there was no fear of it in contaminating your friendship. You know, sometimes you, you plunge into these things and you meet each other in a different way.

Speaker A: Well, we've known each other for over 25 years. I mean, we literally had babies together at the same time. And so once you've gone through those kinds of things, we went through the pandemic together in San Francisco. Uh, you can say pretty much anything and know that this love and trust for each other would carry you through. And believe me, there are times when we were just like, this is driving me nuts. I need to take a break. So. But the fact that we could keep coming back to that foundation of trust and love was what kept us going. And it made the book so much better because we could be frank with each other because m. We could say exactly what we wanted to say.

Speaker B: That's sort of. That's called the psychological safety. That safe space where you can actually be robust and, and fight, if you will. Um, but knowledge that you can, you know, you're both saying it for the good intention. I, I did co. Write a book as well. And, um, it's always complicated because what. What is the voice that you're going to use? How did you go across. How did you sort of finesse that notion of my voice? Your voice, what you do? What I do?

Speaker A: Well, it. It was helped that we. We developed Prof. Professionally in the same place at Forester Research. And so we had a certain way of, uh, a voice that we were used to writing in which was very, um, business straightforward. Uh, no super, first, uh, uh, words, no platitudes, uh, so very direct, uh, very distinct, very matter of fact, yet we could focus on just the meaning of the words. And we had from the very beginning a very clear focus for the book, which was how do you start? Where do you begin? Because this is the question that kept coming up over and over again. There's so many things you can do with AI. Where do we begin? How do we start? How do we create value with AI So what began as much, very much an explanatory book quickly became, uh, a very concrete, structured process to create value. Do this first, do that second was 12 weeks, 12 chapters, 12 steps that you could take. And we just said, there are many, many right ways to create value with AI and there are distinctly some wrong ways to do it. And so we wanted to make sure that we were steering people towards the options, the right options, and staying away from the bad ways that could steer you wrong.

Speaker B: Well, there's no doubt that having a blueprint or some sort of plan, um, to follow is very reassuring, and it helps when you've got so many, so many complex issues. Just to finish on the. The notion of co writing, um, because it was one of the appendices, this idea of how much. Where did you use AI in the writing? And I was one. You know, you say about what AI is good at and what AI is bad at. And it was, it was. Of course, there's what. What do you. What's your intention and everything, but did you use any AI to try to finesse the voice at different times and say, hey, listen, we just plunked in 300,000 or, you know, whatever 300 pages of books of words find, uh, where the tone doesn't always feel good? Or was that ever something you tried to use AI for?

Speaker A: Oh, we tried using AI for writing. Believe me, I personally definitely tried it. Katya, to her credit, was much more skeptical. And I was like, well, let's just give it a try and see what happens. It was just awful. It slowed things down. Um, it's very good at short snippets of writing. For example, I had, uh, described three ethical scenarios that people to go through. And I realized it was much better off, instead of like, saying, this is the way to do it, that we write them as scenarios that could be right or wrong, it could be interpreted any different ways, like a case study. And so rather than laboriously go through it, I just said, just take all the fundamentals that we put in there and make it a case study. Fantastic at doing that. It saved me hours of time rewriting those three examples as case studies. So fantastic at doing that. Uh, it was also very good at creating hypothetical situations. So I'm like, this is the kind of mood I want to set up. Uh, you know, and it was so funny. Katri, like, this is a great example. This is great writing on your part. I'm like, no, AI did that. And there'd be other times it goes, oh, this is clearly a. I wrote this. I'm like, no, I wrote that. So it was necessarily a good thing. I love using EM dashes. Couldn't put them in the book. And so the. Again, we found that it was good at snippets of writing, terrible at holding an idea because of the way it writes, is probabilistic and for the next word. So it can't hold an idea for much longer than a few pages. It becomes repetitive. It doesn't see things. That said, it was fantastic, absolutely fantastic at gathering all the notes, taking our conversations, organizing them, pulling out the threads. Uh, I could engage with our knowledge Base with a conversation. Like, I remember there was some interview that mentioned this, where was that? And because at all of our interviews in a secure location, it could just filter through that. Or I could say find all the examples where security. And it became a really big issue. And we find all those examples so fantastic. From a research, again, just laborious, kind of tedious tasks that are related to writing a piece of work like a book. Um, it was just a lot of information to have to plow through. So just made all of that a lot easier.

Speaker B: One of the things that I'm writing a new book as well, Charlene. And, um, it's not a bad AI but of course for the intellectual exercise and the fun of it, things like a prompt. Hey, uh, have I referred to this particular word, let's say artificial intelligence? Have I used it already? And have I done artificial intelligence, parentheses AI and started the acronym because, you know, sometimes you, you don't know when and how often you might have used AI before. Or you actually showed what AI stands for and, and, and those type of specific prompts allowing you to figure out what is the proper editing what, what. Which, you know, a proper editor does for you. And those are the examples of, of how I've been using it to accompany me. Like you say, it may be more specific tasks.

Speaker A: Right. So we had little scripts that we would run. Uh, for example, we had a list of forbidden words and they would just creep in because I, I tend to write more formally and superiorly. So I would know words like leverage would show up or alignment and capacity. And those are on banned word list. Uh, we just naturally would not use utilize instead of use. It's just, it's just putting more business speak into language. Uh, when it doesn't need to be. It can be much more straightforward. What, what does alignment mean? So we had all the ways that alignment could be used in a different way. So words that could substitute depending on the meaning of that sentence. So we had an AI kind of scrub all of our language over and over again looking for those kinds of words, looking for those places where things just weren't clear because we wanted to be very, very clear. Um, and then one of my favorite things was it would, um, again provide a, uh, counterfactual analysis to us. What else would the audience want to know? We asked it to be that developmental editor because we had one, but then we couldn't work with her anymore. She had some family issues. And we were very effective at editing each other, but we wanted another editor just to take that Perspective from the audience's perspective. What else would they want to know? So we'd go through and just literally edit our things. Like this isn't clear. Tell us more about this. Give an example here. Super helpful in, in making sure we were covering all of our bases. M. Because you get so wrapped up into it.

Speaker B: Oh yeah, totally. I mean the, the whole project of the book. You have two people. Did you say. Did I say it? And, and I mean at the end of the day, ah, this is. For me, the, the. The defining learning for most is just because they told you to do this doesn't mean you have to do it. So retaining that agency, the human element of it and rewriting it or oh, that's a good idea, that's not a good idea. And, and that's also part of using it effectively, presumably in business. I mean really, at the end of the day.

Speaker A: Yeah. I would say here's a draft of a chapter. Give me three ways it can be improved. And, and just, just like, just don't tell me the whole world. But the three most important things that we have to fix in this chapter in order to make it improve. And sometimes we take the advice. Sometimes like, no, that's. We already thought about that. We're not going to do that again. But it was this great to have that other perspective. And then we would give it to another engine and say, tell us the three things that we could do to make this better. Just to get another perspective again. I feel that AI has its good days and bad days. You never know which one it's having, which engine is going to be the best one for that particular task. So I will reprompt things again and again and I will always ask it. Um, ask me any clarifying questions you may have before you start. And depending on the questions I ask, I know that's the one to use because they ask really good questions. They're on it. And I know to use that LLM, um, for that task at that particular moment. But people ask me all the time like, what's the best one to use? I'm like, it depends on the task in a day.

Speaker B: It's funny. You almost think that they might have a, uh, humor. They got slept on the wrong side of the bed, the way you describe it. So, um, we're going to get into some of the elements of the book. But we. What is clear is most people who are listening, ah, are completely aware of the existence of AI. There's. Unless you're living in some dark hole, it is not Possible. So everyone's sort of thinking they've got it, they're reading up about it in Financial Times, they may well be using it personally and probably at some level, you know, everybody's in it. You who are now sort of an, I mean, as you've developed Charlie into this expert with Katya on this, how do you stay up with it? Because it's, as you said at the beginning, everyone can use it. It's everywhere. But how do you actually know that you're on, uh, top of the game? You're not missing out something. It must be one of the continuous challenges. You say curiosity is a great leadership quality, but it also killed the cat.

Speaker A: Right. A part of it is staying really focused on what I want AI to do for me. As long as I say really focus on that, I don't get distracted by whether one model is better than the other or, uh, whether, uh, AGI is going to come in six months or six years. Those things don't matter. I don't pay attention to those things because they don't really impact the work that I do, um, and the problems I'm trying to solve. So my best advice to somebody is don't try to stay on top of everything. It's impossible. And why would you, why on earth would you need to and want to. What's important is to understand what AI can do for you and how you want AI to be helping you and supporting you. So start with the things that you know. You know your problems, you know your business, you know your life, you know your customers, your employees, your suppliers. You know what you know really well, that's what got you to be where you are today. So focus on the things that you know and then pick and choose which areas of AI are important to you. And I say this to board members too. It's like you can sit here and try to boil the ocean with AI, but if you're there to support the development of responsible and ethical AI and to support the well being of the organization to be successful and thriving and to provide shareholder value and stakeholder value, then you should focus on the area of AI that is of great importance to you. So it could be the ethical use of AI, it could be about the future workforce impact of AI uh, whatever is that area that you want to focus on, have historically been curious about. Again, as a board member, you can't focus on everything you know that you are providing a particular perspective that is unique to you. That's why you are on this board. And the same thing as an executive you bring a unique perspective. As a manager, as a leader, as an individual contributor, you bring a particular perspective. So harness the power of AI to solve the problems that you know. And I take a design thinking approach to this. Start with the problem, really understand the problem, and then use AI wisely to help you solve that problem. Again with full knowledge that AI can do some things and can't do other things. So this is about good, uh, prompting good hygiene around those things, but it's also just developing your expertise and fluency in particular with AI. And I think that the meta aspect of this is if you don't know how to use AI, well then ask AI how to use AI because it's kind of like that's how I do it. So I hear about. So my, my thing lately is to develop my knowledge and context graft. Uh, that was developed again I'm kind of inspired by Andre Karpathy's LLM wiki. So I've applied that. I'm using AI to actually create this knowledge graph based wiki. Um, organizations are doing this now. They're creating a knowledge and context layer so that agents can flow easily across the organization. I mean these are sort of cutting edge things, uh, but just sort of staying up on the periphery and applying it to problems that I know I have and helping organizations apply it to problems that they have. So this is sort of cutting edge stuff. It's been out for about six weeks or eight weeks or so, but it's incredibly powerful when you sit down and go, oh, that's a great and easy solution to this. Uh, so again, what are the problems that you have as an organization, as a person and use AI to solve those problems?

Speaker B: I don't sort of check in it, but I believe it's Katya that said, uh, don't ask what AI can do for you. Ask what do you want AI to do for you? As in for you for what? What do you want? What's your need? And then that let that be your hook and energy as spiral. So, all right. You, you. Oh, uh, yeah, I was going to do one more thing which I, I loved, which is, um, when you referred to Erica Goodwin, who is or was this SVP of First Heritage Mortgage. So in a banking environment it comes up with this idea that using AI to spread joy, a profound lesson that sometimes the greatest value AI can create is in making a business more human in these small ways that can scale empathy, can also help you stand out in a crowded market. How so? Maybe you can talk us through that specific idea. Because this idea of AI enhancing humanness, not necessarily humanity, but our, our human interaction, how did that, what was that idea? And how do you think other companies should take this as some kind of lighthouse idea?

Speaker A: I thought you would like that story. So what Erica did at UM is again the mortgage mortgage company. And. But they realized the move is one of the most difficult and potentially traumatic things that you could do. So they started making children's books for the families and they put the child and the into a story of their move from one house to the other. And these are custom made one off books for that family. And because AI makes it so easy to create that they were able to do this for every family that has a mortgage with kids. And you can imagine you've sold the house and you've got the mortgage and you're moving. The last thing you're thinking about is a mortgage company. And here they are showing up with a book. Congratulations on your new home. Here's a book to help your children with the move, presumably with the name

Speaker B: Jimmy, the name of the children, the,

Speaker A: the location, all customize the location, the stories, any other tidbits that the broker can add to it. And uh, it's just like a little, it's a little paragraph that they just fill out and it just creates this book and story for them. So I think. But the thing that really needed to spark this idea, it's a great, you know, customer encouragement and people do referrals for mortgage companies and all the business aspects of it. But it required somebody to think, you know, this is a really hard thing for people to do. How can we make it better for our clients? And this is not something they had to do. That sale is already done. Right. But from a marketing perspective, they go, how do we create that deeper sense of connection? How do we create joy? And I think it's something that we talk about in the very last chapter. We talk about this concept of a superhuman. It's somebody who can use AI very very well. They mastered it, they know how to use it fluently. And instead of just using that to do more work, because you can always just do more with it, they use that capacity, they use that time, they use these new abilities to deepen what is truly unique that makes them human. So deepening their empathy, their ability to do self reflection, their sense of intuition, uh, judgment, and finally wisdom. I mean there are many other aspects. We just centered on those five things. Uh, AI could emulate empathy. In fact, it can probably do a better job of Being empathetic, but it cannot feel empathy because it has never felt pain 100%. Yeah. So it's. And so these, these are things that are human. And when we think about AI augmenting us as humans and our capability doing these things, it can help us deepen those areas of humanity. And the reason why we think this is a leadership issue, not a people development issue, but a leadership issue is you have a choice in your organization. How will you use that extra found capacity? Will you just do more work and create value with that which is great, or will you invest that and the capital v value of people in their humanity? Because in the end, if you have an organization filled with superhumans who are able to, to again just leverage AI, use AI in the best way possible, but can also develop their humanity, you, you develop a really unique aspect of the organization that is a competitive advantage, I believe. And so you have to intentionally strategically invest in that, reinvest that into your people. And it's something. We just don't invest in people in their development. We can talk about leadership development and everything, but fundamentally we don't do a lot of that and we do it out of necessity when there are problems. Uh, but if you were to proactively invest in people being better humans, what could your organization be? Be more empathetic to your customers and to each other. Be more intuitive in the way we make decisions and judgments. Uh, be more self reflective to understand what are the impacts of our actions and our thoughts and our words on other people. So again, it just adds to a very interesting dynamic where it's AI is augmenting us in our humanity and also we are injecting our humanity into AI because it is not human. It doesn't have these features and allows AI to be in greater acknowledgment of these aspects. I recently was talking to somebody and they put into their instructions for chatgpt their values. And it said as you're making these judgments, as you're making decisions, take into account my values. And I thought that was such an interesting way to inject humanity into the AI, into literally the instructions, it's global instructions of how it should work. That's truly living your values and having your AI live your values alongside with you.

Speaker B: So that brings up a whole lots of thoughts for me. Two zones I want to go into. The first is so great idea, mortgage stress, uh, children's book. At some level that's an idea that anyone could copy because I oh, great idea, I'll do that and you do you know, fast followers that you talk about that don't be a fast follower, you become a slow loser. But, um, at some level, the bigger story for me has always been about what is unique, proprietary, about what I do and how I do it. So the link of who I am, I, what are my values? And the idea of library joy needs to be completely congruent with what you have been and what you intend to be. Because if you're doing that on the back of some sort of, that sort of sharky private equity company may not sort of land quite as well. Uh, anyway, so then all of this has to live with some sort of tension with regard to transparency and proprietariness. The IP you talk about IP of the prompt, um, there's a need to have transparency. But how does transparency and IP coexist? And maybe to finish my monologue at this point, I've long thought that your IP relative to your AI should be the sum total of all your interactions that are specific to you and your company, you and your employees, all your written documents and everything. And yet within that, you may have some trade secrets, some things you don't want out in the public domain. You know, think of that, uh, Hollywood film company, things that we don't want to have. But so anyway, how, uh, do you fix that tension, Charlene, between the, the need for transparency, top of your pyramid that you talk about, and the idea of IP and stuff that you don't want out in the open?

Speaker A: Well, I think transparency, again. I wrote a book called Open Leadership that asked the question, how open do you need to be in order to build trust? And, and, and it's, it's not about being 100 open, because nobody ever is 100% open. And I don't think anybody want to go near you. So they knew everything about you.

Speaker B: Yeah, but by the way, I don't even know all. I don't even know everything about myself as it is. So it's bar me to think I had somebody on my show one time, I said, and she said, it takes more than a lifetime to get to know yourself. So you're being transparent about what?

Speaker A: Yeah, about all of my things that you. Again, you. I, uh, don't even understand. Right. So given that, what, what does transparency mean? What needs to be transparent? And a lot of it is about how you make decisions in an organization. And how does AI impact the way you make decisions? Uh, let's just look at it from the perspective of employees. Employees just want two things from their employers. They want you to Be honest. And they want you to be fair. Just be honest with me, don't lie to me, and then treat me with the same respect and treat us all equally in whatever way you define fairness. And that in itself is an issue because are, uh, we talking about fairness of equality of outcomes or equality of opportunities? Those are two very, very different things. So even our definition of fairness is up for debate. So given that again, and transparency is, what will you be transparent about and what will you not be transparent about? For example, you may not want to be transparent about people's salaries because of whatever reasons other organizations may want that all out in the open because that's, they believe in that. Uh, you may be transparent about, uh, the things that are happening from a legal perspective. Lawsuits, um, your, um, plans for how much you're going to pay for your next rent in the next year, all those negotiations, or you may not choose to be. So, uh, what needs to be out there? When does it need to be transparent? When do you pull it back? And being transparent about what you will be transparent about is how a big part of this battle. And so the reason why transparency at the very top is you need to have, and this is at the top of an AI trust pyramid that we built. At the bottom is safety, security and privacy. Those are just foundational things. You have to have those things clearly protected. Fairness, as we discussed, uh, reliability, which is accuracy and quality, responsibility and accountability, and then transparency. The reason transparency is at the top, because they have to have the other things worked out to know what you will be transparent about. And you need transparency to build trust. So your, I, I'm not saying that you have to be completely transparent about all the IP and the prompts that you use, because that's your secret sauce of how you get work done. And even more importantly, what's really a secret sauce are, uh, the people using those prompts, the people using the proprietary things, the way they apply it, how they change it and adapt it and improve on it. That's your true secret sauce. Are the people using these tools, uh, but the tools, the decisions themselves. The question becomes, how are you using those tools to make decisions? If you're a mortgage company, how are you approving or not approving the loan? When are you using AI? If you're hiring people, when are you using AI to screen through resumes? How are you ensuring that there's no bias? And all of this, again, the fairness issue. So these are, these are the questions that you have to be transparent about. And, uh, being very clear about how you are looking for bias, how you are adjusting for bias, where you believe it exists, because there's always bias in any tool that you use. So what biases are you trying to equalize against? So these are the ways that you can be transparent again, to build trust and the way that you do work with people, the way that you make decisions,

Speaker C: all.

Speaker B: Ah, right. So one of the, um, really interesting cases you anonymously write about company A and company B, uh, and company A is sort of, let's go for it adventure, explore, test and learn. And then, oops, fuck up, excuse the French. And then company B is, no, no, no, we need to be very careful. Regulations, uh, guardrails, governance, let's do all that. And by the end of six months, they've done zippo in. In the, in the realm of that tension, to the extent that I want to create a proprietary IP like AI, to what extent do we have to worry about the public domain of if I put in all my emails and my, everything to get my, my culture, my voice, my values, as they are de facto ex, you know, made to, uh, come alive, how do you square that box or that circle when it comes to figuring out confidentiality, security and yet experimentation and getting everybody to have fun with it?

Speaker A: Right, so there's a question about if I use AI, is it safe? Is it going to keep my secrets safe? And there are definite steps from a security perspective. In the same way that we put our most confidential information into the cloud, the same security protocols exist. Uh, one of my favorite services, uh, out there is aws, Amazon Web Services. Many organizations have their data already up there. Uh, and they have a program called Bedrock, which is we can apply AI on top of your data and it's secured, it's only within your data follows the same security parameters. You literally push a button and you have all these different models available to you. And it's one of the easiest, most secure ways I can think of right now. If you're already using aws, it's a fantastic stepping stone because it requires almost no technical skill to enable that and it's just immediately available. Uh, again, if you're a bank, you probably wouldn't do that. You would build your own secure instance of AI that's completely knocked off from the rest of the world. I mean, there are different levels of security. So from my perspective, worrying about your IP escaping into the wild is much less of an issue today than it was maybe even a year ago. There are definitely. I mean, again, everything I use is locked down as much as I can. Um, and I'm trusting, and this is the key thing, I'm trusting that these AI models say that they won't use my data to train their models. And I'm trusting that they don't do that. Uh, and if they were to ever violate that, that would just be the end of their business if they were to do that. So that's the level of trust that I have. That said, I still don't upload the most proprietary information for myself of my clients. I make sure it's clear, clear of what I am uploading, what I'm not have permission to do that. So it's again, just being super careful about what you put into these engines or not. But for the most part, understanding, you know, your voice, your knowledge. Again, the ideas in our book are not so proprietary that even if it were, and I know this as you know this as an author, your words as soon as you publish them are out there, they get scanned, they get ripped and there's nothing you can do to prevent it. You can about all the cease and disorders in the world. Nothing can prevent it. So I'm very comfortable with it because if somebody can, um, again, take those ideas and I'm talking about them all the time and use them, uh, but I don't necessarily get credits. It's better that it's out there in the world than not. I get more than enough credit coming back to me, um, just by the volume of content that I create out there. So, you know, one of the things is, you know, being comfortable with creating IP and knowing what happens to it. Um, but I think to your company A company B scenario, this is what I'm seeing right now, and I'll give it a name, it's called the AI Hesitancy gap. I know it's going to be transformative. I'm still experimenting with it. I'm dabbling in it. I'm not completely comfortable. I'm not 100% sure it's going to be safe. That water looks really dark and murky. I don't want to jump in. So I call it fomo Fear of missing out and fogy fear of getting in, which is pulled in both of these directions. Right.

Speaker B: Are you a fogey? Are you an old fogey?

Speaker A: Well, this is a fogey question. And what's really holding you back? And oftentimes it is the sense that the perceived value is not as great as the perceived risk and especially reputational risk that I could be wrong. And so company A being really to just like jump in, experiment, fuck up basically. And like, okay, well that was wrong. Let's keep going. And I think Clarina, uh, the credit company is a great example of this. In 2023, I think they laid off like 800 people and said, we're going to be the biggest users of AI. We're the guinea pigs for OpenAI. We just want to do all of this. And a year later they go, that was too much. Our customer, uh, satisfaction scores fell through the, through the ground. Uh, we have to hire back these people and uh, we push a little bit too far too soon. And again, I think they lacked some of that empathy that was needed early on. But the CEO says, yeah, we mess up. And my goodness, thank goodness we did. But at least we tried. And we learned so much more. And we are now so far ahead in terms of using AI, learning more about what AI is good, when humans are good. Yeah, we look back, we wish we hadn't done it that way, but now we've learned a lot more and that you're better off in that way. And they had the capacity and the resilience as a company and as a leader to do this compared to Company B who's so like, um, I don't know, should we do call summarization? What happens to this? Like the most basic areas of AI, they're still hesitating around and it's out of fear that's overwhelming the opportunities, the benefits. And I think again, that weighing of these two things is what's really going to set apart the people who are sort of dabbling with AI, maybe even adopting it. But they haven't gotten to the place where they're adapting and changing and transforming their organizations in the way that's needed to have AI really take hold.

Speaker B: So you who have spent so much time doing transformation, it strikes me that one of the biggest needs is to know how to learn from our failures. And yet running at 1,000 miles an hour in an organization where if I admit that I screwed up, that could be a, uh, bad idea for my career. The ability for uh, you to set down guardrails for failure and learning from failure, exposing what I did wrong and having that sort of safety to be able to say that, I feel like that's one of the biggest issues. That if somebody wants to be a company A, they have to know how to have that sort of second step allowed.

Speaker A: Right. And I used to be, again, uh, in my book Open Leadership, I wrote a whole entire chapter, um, called the Failure Imperative. And I've changed my thinking around this a bit because everybody hates the idea of failure. I don't care who you are. Like celebrating failure, running towards failure. It's like, it's just not a natural thing to do. So I think about it in two ways. First of all, what if we were to think about this as learning, and you're setting yourself up for the psychological safety to learn. And in order to learn, you have to figure out what works and what doesn't work. And I look at it this way. In order for a project to go forward with AI, do you need to be 100% sure that you're going to get the results that you imagine it to be 100% sure, or can you be 80% sure? Can you be 60% sure? How sure do you need to be before you pull the trigger and say yes and know that I may get halfway there? And the rest of it I'm going to have to figure out. I mean, I have all the answers. Um, and I focus on this one idea of minimally viable data that you need to make a decision. And how do you make decisions around things that are small enough that you can practice developing that judgment that we talked about before to be able to say, yeah, 60% is right for this, but I really need more like 80, 90% for this type of decision and knowing what types of decisions you can easily make and be okay with less than 100%, uh, outcomes. Being perfect is a key part of leadership. Because if you are 100% sure every time, that's not leadership, that's management. That's managing the status quo, where you know everything's going to turn out exactly the same. Leadership is when you lead into the unknown, when you lead change. That's why we need leaders and we need brave and courageous leaders. Because it's not 100% sure. So organizations that are capable of building this scaffolding, the guardrails. I'm going to say the G word. The governance. The governance around this. Governance is one of my favorite words when it comes to transformation. And I'm not talking about good governance. Uh, Katja has this beautiful saying. Structure without flexibility is bureaucracy. Flexibility without structure is chaos.

Speaker B: Chaos, yeah, I remember.

Speaker A: Yes. You need what we call Goldilocks governance. Just the right amount to ensure you can go as fast as you can safely, but also really make it uncomfortable for you to just stay still. So good governance clears the way. It allows you to say, these are the places not to go. These are the places that are safe. Defines the edges really clearly. So what you do when you have good governance, very clear governance, just right, go deluxe governance, you can take your people to the edge of what is possible. You show them this is where we can go, don't go past that, but also stay at this edge, stay out here. Uh, if you've ever been to a playground with a fence, guess where the kids are. They're at the fence because they're going to push the boundaries, see how far they can go. They know that as long as they stay inside the fence, they're going to be fine. They can explore anywhere inside of that fenced area. So uh, if there's no fence, they stay close to the playground structure. It's a playground paradox. And we need to understand that, that good fences, good guardrails, good breaks, allow you to go fast, allow you to explore. So the, the challenge I think for organizations, for leaders, is to find, is to spend time defining what you can do. And it's the same idea as you can be open, but how open do you want to be? What will you be open about? When you are clear about what you, you can be open about what you can explore, then you allow that exploration to happen. But it's a lot harder, it's a lot more work to define that than to just say simply, no, don't use AI. And it's actually more dangerous to say that because guess what? People are using AI. The amount of shadow AI that's out there is just remarkable. And all you have to do is just say um, to the it, just pull all your IP records and to see what, what size people are actually accessing through your networks. They may be doing it on devices, but they're still doing your network. And guess what? They'll go those, they're using AI constantly so you can try to ban it. You're much better off controlling it and governing it than turning a blind eye and hoping that it doesn't exist, because it does.

Speaker B: It strikes me, Charlene, in your 90 day approach, in the beginning, you like week two, I think you talk about governance and that's such a key part. You also say, don't believe in AI strategy, just have a strategy. It strikes me in that when you try to put together actually what is your strategy that everyone understands and follows, who are you? What are your values and what do you mean by you say we are family, what does that actually mean? That's a conversation. And when you talk about having a governance, that's fair and no bias. These types of works are profound and heavy and hard to do. In quick fashion, unless you already have done some pre prep work, I would say otherwise, it gets difficult. Um, and I just, I want to get to one last question, but as you were speaking about the boundary, it reminds me of my time at Burning man, where of course, I had to go out to the perimeter where there are no stars out there on the perimeter, as Jim Morrison said. But last, um, question, really. And, and this is for people who are in work and we talked about, you know, staying up with Joneses, staying up with AI and this idea of productivity and saving time. But it, it just feels, Charlene, that as much as people are using AI, I don't see anybody going back and kicking back with two extra hours every day. It does feel like we're moving towards overload, burnout. You talked about these superhumans, but how

Speaker C: do you

Speaker B: advise people to figure out how to actually save time or not burn out anyway from the speed with which we're operating, the amount of stuff we have to learn every day and transform everywhere, and the dangers from left and right and outside and inside your industry and inside your company.

Speaker A: When I work with leaders, I ask to see their calendars. Because your calendar exposes what you value. Where do you spend your time? It's a simple question. Where you spend your time is what you value. So do you value meeting after meeting after meeting, where you just meeting with everyone with no time to work? Do you value setting aside a few hours a week to think? Do you set aside that time to learn? I was speaking with a leader this past week, and she was saying, you know, I've done this before. I've done this, you know, where I had to learn how to use social media, I had to learn how to use mobile, I had to learn how to use, you know, Slack and all these other tools and whatever it is that I had to learn. I always set aside time on my calendar and I got an accountability buddy. Um, and I'm doing that right now. I'm putting out time on my calendar to do this on a regular basis, to get fluent with AI because there's no way I can ask my organization to become fluent with AI if I am not. I cannot lead an organization where I am wandering in the dark and unsure about things. So I need to get up to speed on this. And again, I don't think there's any way to escape it. And when you look and say in all of that frenzy, of all that overwhelm what is most important to you and your organization, um, that is strategy. Strategy is what we will do and what you want to achieve your goals. And if you're just on that hamster wheel, that is not strategy. It's the same way when you have a long list of use cases for AI, that is not a strategy. A, uh, strategy is intentional to say, what are our objectives, what are we trying to achieve, what's the future that we believe in, and how are we going to get there? How we're going to get there is that strategy. And AI is just a piece of that. It's just a tool to enable that strategy. It doesn't deserve its own strategy that's separate and independent of your strategy. It's like saying, we're going to have an Internet strategy, we're going to have an electricity strategy. It makes no sense. It is a tool to help us achieve that strategy. So let's just keep things in perspective. What is the most important things that we're trying to do here? And so when somebody's overwhelmed, I'm like, let's take a deep breath. What's most important to you? Because if we can center on that, then we can figure out how we're going to use AI. To Cacio's point, what do you want AI to do for you and keep that in perspective. It is just a tool, and you have to learn how to master the tool.

Speaker B: Love it. I was just trying to look through my notes. You did say it's possibly a good idea to have a specific AI person in your executive team, did you not?

Speaker A: Yes, I did. Uh, we wrote about the fact that. And again, we refine this a bit more, too. There was a great HBR article in January that talked about who owns AI. And the reality is multiple people in the organization are going to own a piece of the AI. And so it doesn't make sense for one person to quote, own AI the way you think about. You need one person who owns the outcomes of AI to make sure that everyone is coordinated. I think of this person as a conductor or an orchestrator who makes sure that, again, uh, from a conductor's perspective, that the score, the strategy is clear to everyone and that we're all playing off the same score, the same strategy, not individual department or functions or business unit strategies, that it's all unified against our overall strategic objectives and that the tempo is being set. We're going to do this first, this second, this third, this fourth. And everyone understands this is a tempo. And if you're falling behind, we're going to give you extra time and resources and say, what's going on here. Uh, why are you falling behind? Because you've agreed that this is a roadmap. We need you to be pulling in the same direction at the same speed. So you need one person who wakes up every single day whose job is to make sure we're driving the valued outcomes from AI that we expect. Because otherwise, if it's everybody's job, it's nobody's job to do that.

Speaker B: Totally. Hence so many transformation processes fail. I was listening to Aravin Surinas CEO co founder, Perplexity, and he. He also talks about this conductor idea anyway, and the idea of, of focusing on the outcomes, the outp. So brilliant insight, Charlene. I wish we had a lot more time. Uh, because I didn't even get through, I don't know, 90% of the questions I thought I would. I just followed you down rabbit holes. How can someone track you down? Get books, follow your readings. Um, get. Hire you as a speaker.

Speaker A: Yes, thank you. Um, you can follow me@charlenelee.com. uh, and the book is at winning with AI book.com and I'm all over LinkedIn. I just started a substack too as well. So, um, speaking, I'm running workshops for organizations, so would love to be able to engage with people more just in any of these channels to learn more about what you all are doing.

Speaker B: Spectacular. And hopefully we will have a chance to catch up. When you next come to London, England, visit me in West Kennington. Charlene, thank you so much.

Speaker A: Thank you.

Speaker B: So a really heartfelt thanks for listening to this episode of the Mentor Dialogue podcast. If you like the show, please remember to subscribe on your favorite podcast service. As ever, ratings and reviews are the real currency of podcasts. And if you're really inspired, I'm accepting donations on patreon.com enterdial you'll find the show notes with over 2,100 blog posts on minterdial.com on topics ranging from leadership to branding, tech and marketing tips. Check out my documentary film and books, including the Last 1, the second edition of Heart Official Putting Heart into Business and Artificial Intelligence that came out in April 2023. And um, to finish, here's a song I wrote with Stephanie Singer, A convinced.

Speaker C: I like the feel of a stranger tucked around me Precipitating the danger to feel free Trust is the reason Still I won't to the lie I sit here passively Hope for your respect Anticipating the thrill of your intellect maybe I tell myself there's no use in me lying I'm a convinced man building an urge I'M a convinced man to live and die subvert A convinced convinced man in the arms of a woman I'm a convinced man Challenge my fate I'm a convinced man Competitions in name A convinced man in the arms of a woman despite revenges and struggle with deceit Live for the challenge so life's not incomplete what's wrong with challenge? I know soon we all die I like the feel of a stranger tucked around me Precipitating the danger to feel free Trust in my reach and let me show you why I'm a convinced man Practicing my lines I'm a convinced man Hearing these gun finds A convinced man in the arms of a woman I'm a convinced man Put me to the test I'm a convinced man I'm ready far and around convinced man in the arms of a woman mhm. Sam I am convince you convince me baby. I'm a convinced man I'm a convinced man I'm a convinced man in the arms of a wife Mama I'm a convinced man so convinced yeah I'm so convinced mhm so convinced so convinced so convenience I'm so convinced I'm convinced.

Speaker B: M have you ever wondered why songs on the radio are popular? Why do certain movies get made even though the premise seems completely random? Why are concert tickets costing you $3,000 but nobody makes any money touring? Well, on my podcast, Breaking down the Biz, we answer all those questions and more. I'm Seth Schachner. I have over two decades of experience in the entertainment and the music industry. And every week I talk to insiders that lend insight and expertise on the media you know and love past, present and future. Subscribe now on your favorite podcasting platform or watch us on YouTube so you never miss a beat. Let's make sense of this industry together.