How to Turn AI Into Real Business Value | Liat Ben-Zur

Everyone's going to have powerful models and there's so many powerful models to choose from. And there's so many vertical AI solutions that solve very specific problems that all of us could use in different companies. Your advantage is going to come from what sits around the model, your proprietary context, your trusted customer relationships, your understanding of your unique workflows, your better feedback loops, your faster learning, your judgment to know what should remain human. You don't win by following the playbook. You win by rewriting it. 700 episodes deep with the people who actually built something real. No theory, no fluff, no shortcuts. This is Right About Now with Ryan Alford. Leon, welcome to the show. How are you? I'm doing great. Thanks for having me, Ryan. Great to be here. Yeah, I appreciate you joining. I am always ready to talk about AI. And also, it forces me to look in the mirror on, hopefully it's not brittle. You want to be a leader and good leadership with everything going on. It made me think about a lot of different topics. So I was excited to talk to you. I'm excited to talk to you too. There's a lot to talk about. Let's set the table just really quickly with your experience leading up to the book. I've been an operator in tech for about 30 years now. I spent almost 20 years at Qualcomm helping to drive the transformation from 2G to 3G to 4G to 5G. We were powering most of the phones that are out in the world. lead the transformation towards the internet of things. So how do you get connectivity into everything else beyond phones? I was really active in defining kind of the open source software strategy around that. Worked five years in Europe at Philips where I helped lead digital transformation around personal health. So helping to connect devices from the hospital to the home. How does that change business models? How does it change the engagement with customers? How does that change how we can manage and monitor? health the last five years here in Seattle, which is where I'm calling in from now with Microsoft. And I led the two largest consumer businesses, Microsoft 365 subscription business and the edge in being P&L. Really kind of been involved in technology in a lot of different ways. I left Microsoft about three years ago and started an advisory firm around AI. I was lucky enough while I was at Microsoft to help launch the first AI GPT-4 commercial product with Bing. This was actually even before OpenAI launched GPT-4 themselves. It was very, very interesting. And quickly, I realized that this was going to transform every single industry and every single business. And I wanted to go out and help as many of those as I could. Been on a really exciting journey, just helping companies across different sectors, healthcare, finance, automotive, industrial, retail, foods. What does AI mean in those industries? How can CEOs take advantage of them? What does successful transformation look like? That's kind of the journey that I've been on. From that work sprung this book. I worked in wireless on the ad agency side for 13 years. I worked on other clients too, but it was like 75% of what I did was working with Verizon Wireless. From can you hear me now to 5G to everything else in between. And so I started reminiscing thinking, and I don't know, you've probably done this yourself. You've worked so deep in these companies at corporate level and seen it all. What would we have done with AI in 2007? That's a great question. So many things. And I make this analogy all the time. It's like before we had the calculator, before we had the microwave, that's a very simple rudimentary. I don't even know if that does it justice completely. Before we dive into specifically how things can go south or how time could get wasted or how bad decisions might could get made. Talk to me about just your opinion on how transformational this AI is going to be and already is. I've been on the leading edge of digital transformations for the past 30 years. Nothing compares. Not when we first even moved into the internet, when we started doing wireless, when we went to mobile, we went to cloud, we went to the internet of things. Like all of those are huge and they've completely transformed society in so many ways. But the speed at which this is happening, it doesn't compare. And the speed at which it impacts every single industry. Even when we went to the internet, I remember how it was slowly kind of getting into all the different industries. Everyone started making a website, but it was relatively slow. Within three years, I went from talking about AI between Microsoft and OpenAI, which is the bleeding edge. You would expect those guys to be talking about it to working with every single non-tech native industry. That has nothing to do with this. They're all deep in it. They're all trying to figure out how to transform. So it is it's just part and parcel of every single business, every single function. Every function is touched by this, whether you're in legal, finance, marketing, IT, ads, customer support, product. If you're not using this, you're going to be left behind. It's very transformational. Isn't it the equivalent of sitting next to someone that's using one and you're not? If you're taking a math test and someone's got a calculator and someone and you're allowed to use it, isn't it that big of a leap, if not more? Yeah, I think it's probably bigger than that. Anyone can use it, but how do you use it well and what does good leadership look like? And that's now we start getting into some of the stuff that I write in my book, The Bias Advantage. Are we still writing the book of what you just said, what it's best used for? That feels like that chapter hasn't been fully written yet. There's a lot of writing out there on use cases and how to use it and what is the technology and why should you use the technology? And I think there's... never endless amount of writing out there around that. What I hadn't seen anyone talking about or writing about and what I started observing in my experience working with a lot of these companies and a lot of the leadership teams at these companies who are trying to figure it out is not just how does it change how you work, but how does it change leadership and power? And that's really what I focused on in my book. When I think of this, and not that my imagination matters more than your real world experience, but a lot of everybody's doing it. How are we doing it? We've got to cut costs because AI is cutting costs. Everything's like the race towards those things, whatever they might be. It's like fire aim ready. three years ago when I started on this journey, that was a conversation. A lot of it started with FOMO. It was either boards pressuring leadership teams. Why aren't you using AI? What are you doing AI? Or leadership teams telling their organizations, Hey, we don't want to be left behind. Our competition's using AI. What are we doing? And I think what that ended up doing, and we're seeing a lot of it today, there's a bunch of research out. There's a whole bunch of AI initiatives are actually failing to create business values because a lot of those teams started with what can AI do? are all the ways you guys can be using AI to try to save us money or save hours or whatnot? And instead, they should have been starting with what's our most valuable business problem to solve? Where are our decisions too slow? Where is our cost too high? Where are our customers really frustrated? If you don't anchor AI to like real measurable operational problems, metrics that you can track, whether it's acquisition costs or throughput or retention or whatever, really what we're learning and you're seeing it now is you have a bunch of companies that are just burning through compute. It's really important not to confuse AI activity with AI transformation. Just because you're using AI doesn't mean you're actually moving the P&L, impacting the business and driving transformation. I think in the early days, there was definitely a lot of that that was driven just from FOMO. And then the other thing that I'll say is that like today, most of the AI projects that are not delivering impact are not failing because of technology issues. They're not failing because they're not using the right tech stack or they picked the wrong partner because there's so many tech stacks and there's so many partners you can use and they can all deliver impact. They're actually. failing because the organization just can't absorb the change. That could be cultural. It could be fear. It could be leadership. It could be that companies are just trying to buy a bunch of technology without redesigning the work around it. But if you just go leverage AI and you stick it into your org without changing anything in the org as is, it's like you're buying a Ferrari engine and trying to drop it into a lawnmower. It's going to be super powerful and super noisy, but it's not going to take you anywhere. And so you really have to rethink a lot of organizational stuff in order to take advantage of this. It's my own experience building companies and getting leaner and being a practitioner. The burning compute both made me nod and made me want to fall over on the floor for how much I've done that myself. I can see at scale how big of a problem that could be because it's usually operator error or just compute with no end. You think you have the end in mind, but you haven't completed that picture. It's almost like getting in the car and hoping it takes you to the destination. clearly on your metrics on the goal and you got to be tracking it religiously. Are we moving it or not moving it? Can we measure it or not? Otherwise you're just moving the goalpost and you're really not going anywhere. I've moved that goalpost so many times. I have too. And I think that's part of experimenting with AI, right? Anyone who's on the forefront is also experimenting and you're going to be wasting a lot of compute when you do that. One time I was kicking a 10 yard field goal. The next time it was a 90 yards and then it was outside the stadium and I was trying to kick it. And I'm like, and I didn't even know I was outside the stadium, you know, right? And if anyone has experimented with AI, they're getting all these analogies, and most people probably have. I'm talking about Liat Ben-Zor. She is the CEO of LBZ Advisory. Liat, where do we go brittle? Where does leadership start falling down? What are the core tenets or what's the signs, the signals of that happening? The big mistakes that I'm seeing executives make when they invest in AI is they delegate transformation. They hire some chief AI officer, they'll create some council or something, and they basically say, go figure this AI thing out. But what happens when you do that is it typically lets all your functional leaders a bit off the hook to continue to run the way they've always run while someone else is experimenting with AI around the edges. I really believe, and from what I've seen in successful transformations, each of your functional leaders need to feel ownership. The head of HR needs to own how recruiting, performance management, employee support have to change with AI. Your CFO has to own what happens to forecasting, to controls, to working capital, to accounting, leveraging AI. They need to work with the AI vendors that are really supporting finance. Your sales leader has to rethink prospecting and pricing and proposals. The technology team, they can provide your guardrails, your platform, your data security, but they cannot own the data, the business transformation for everyone else because it is not, this is not just like an IT cyber system. It is transforming and forcing business leaders to rethink their own workflows. Massive, massive, massive change that I see a lot of leaders struggle with. And then the other mistake is just measuring the wrong things. Kind of like what you and I were just talking about. There's a lot of teams and leadership executives that are asking how much code was written by AI, how many users are using this new tool or this chat bot, how many hours have been saved. And they're looking at these metrics that all look fantastic and they're giving each other high fives. That's activity. That doesn't necessarily mean business impact. So, you know, I want to look at, did our cycle time fall? Did our conversions improve? Did our claims leakage decrease? Saving five hours of Joey's time is great, but what do you do with those five hours? Does the company eliminate costs? Do we handle more volume? Do we make better decisions? And so really measuring the right things is massive. Who's getting it right and who's getting it wrong? If you want to go there, I don't know if you want to name names. You just define what maybe success starts to look like and what becomes the true best measure of that. What you just said is it makes a lot of sense. As we all know, the road to hell is paved with good intentions. And sometimes it gets confusing with knowing what's the right intent and the activities and they feel like you're getting somewhere. What does a successful setup look like that enables what you just described? The companies that are getting it right, they're kind of doing four things differently. They're definitely starting with a very crisp definition of what's the most painful, valuable business problem we need to go solve rather than like, here's some sexy new tool I saw on Twitter on X. Can you guys show me what we're doing with it? The second thing is redesigning their whole workflows. They're not just saying, Hey man, I want you to figure out how to accelerate what you've been doing for the past five years and get it done faster. They're not just bolting on an AI assistant onto 12 broken steps and calling that transformation. the question, why are we doing it the way we've been doing it? Do we need to be doing it this way? Do you need this team to pass it to that team to pass it to that team to then get reviewed by this team? Or can we just completely change the entire flow because of AI? Third is really accountability. They're being really clear on who's on the hook for results, not just like an innovation group, committees. So accountability is a big one, especially with AI, ownership and accountability. And then I would say the last thing that these teams do really well is they build the guardrails along the deployment. They don't kind of wait until they build all this stuff, see who they've hurt, what they've messed up, and then go, oh, okay, let's add some security. They think about security and privacy and evals and human review and accountability. And they design that into the workflow early instead of waiting until something goes wrong. I tend to see these as characteristics of teams that are doing it right. And generally speaking, the winning companies are also much more willing to stop things because you can't be in this AI game without experimenting. You're going to have to do some experimenting and you want to encourage experimentation, but experiments don't all deserve to scale. So you have to be, again, be really clear on establishing the evidence required to keep investing and then boom, make those tough decisions. We're going to stop A, D and F and we're going to double down on B and C, getting real clear. Sometimes winning the game is stopping doing something. You always think it's one more thing you do. Sometimes it's stopping something that's hurting you. And I think that is critical in this. It's gotten real easy. It's almost addictive because of how good it is. For me, it's almost become like watching TV because I like the way my brain works. It's almost addictive using it. But then it's like, okay, instead of going down like the TikTok rabbit hole of watching useless, I'm going down like the compute. Man, it costs money. I can see people doing that. They feel like they're solving it, but they're not knowing when to ditch it when it's not getting to where it's solving a true business outcome. And that's what it is, outcomes. Outcomes matter more than activities. How are you working with teams? And when you go in, what's the first sign of we need help? I'll tie this a little bit to the book that I just launched. The book is called The Bias Advantage. And the whole idea behind that is that AI is really changing what we need out of leadership. A lot of people, again, are just talking about how AI is changing how we work, but it's really changing leadership. What I'm seeing is that when answers become cheap and anyone has expertise and knowledge at their fingertips with AI, the best leadership shifts to really knowing which questions matter, which answers deserve to spend deeper analysis. They really understand trust because when it comes to machines and AI, we have a pretty big trust issue. issue around that. They think about what data is missing. AI is only as good as the data that you train it. If there's data missing, then you're going to have problems with what comes out. And they think about who's going to take responsibility for the decision. When something goes wrong, you don't blame a machine. You got to have accountability. What this means is leaders really need to have strong ethical clarity, adaptability, and really the courage to just question a system that appears to be working. Right. Hey, I'm doing well. I've been doing well for 20 years. Why do I need to rethink my workflows? Why do I need to rethink how we do things? It seems to be working. And the argument that I make in the book is that many unconventional leaders have to develop a lot of these skills that I'm talking about early because the conventional system didn't automatically work for them. And so the book really asks this provocative question of what if the qualities that a lot of organizations dismissed or penalized or just fail to see are exactly the qualities leadership in this AI era now requires? What I'm seeing right now in a lot of these companies is that AI is really... testing and exposing leadership. It exposes brittle decision-making. It exposes outdated talent models, unclear accountability. It exposes cultures that reward confidence over judgment. Like a lot of stuff is coming to the surface right now because of the changes from AI. That's a big premise of the book. And it's also a lot of the conversations that I'm having with companies. Is the world of AI that we're in, in the corporate environment that you're describing, and is creative leadership valuable or invaluable in a world where the numbers are so apparent? When it comes to like data and insight, if AI knows more than all of us, what do you need to do to stand out as a leader if you're going to get all the answers anyway, if the expertise is already in the machine? I'm just saying, if you could paint by numbers every time, if you do it right, So if you do it right, a leader defines the problem. A leader decides what trade-offs are acceptable. A leader understands the context that's missing in the conversation. A leader takes accountability when the output of a system can cause harm or when the strategy fails that you're rolling out. many ways, AI is going to be better at us at crunching the numbers and making a bunch of recommendations. It can recommend that you close a facility, but it's not going to look at the affected community and the employees and I and own the choice. It can generate a beautiful strategy, but it's not going to bring teams along with it to go execute that strategy. It can can give you a bunch of great answers, but leadership is deciding what is the right answer for this moment, knowing that there's a bunch of relationship tensions that are in the room that you understand and the AI doesn't understand. And so the value of leadership is moving away from being the person with the best answer, moving away from being the expert with all the numbers, and it's moving towards being the person with the strongest judgment, with the most clear ethical roadmap, vision, the ability to sense and smell bias before the machine does. AI has moved the needle. In a lot of ways, it's made everyone a B player because everybody has the answers. But it's also changed what it is to be an A player. And that's an A player is a leader. An A player is whatever you do. It demands a side of us, whether it's creative, whether it's vision. Vision is what I keep coming back to, especially when I was looking through the book and like what you just said. This is vision because vision is seeing around the corner. Vision is understanding unintended consequences. That's one of my favorite words in leadership. Yeah. He who understands actually the other side of what happens when a decision gets made ahead of others. That is vision and understanding what's around the corner that's not always so obvious. I couldn't have said it better myself, Ryan. And a lot of that is a premise for why I've observed that unconventional leaders tend to have an edge in doing exactly that. Because when you've spent your career reading two realities at once, there's a formal reality. It's what the company says, the org chart, the state of policies, the cultural norms, expectations, all the formal rules. But then there's like the lived reality, your operating experience, who in the room actually gets trusted, who gets rules that are selectively, if you will, enforced, who gets the exceptions, who doesn't, who gets listened to. That dual awareness is enormously valuable in an AI driven organization because AI is great at finding all these patterns. But the leader has to notice what's absent and whose experience never shows up in the People who've had the luxury of trusting the system tend to not always look for these signals. But those who've kind of been on the downside of bias, they tend to be very good at seeing where things break and smelling things before it shows up in the data. By the time it shows up in the data in an AI world, it's probably scaled. And that's a very dangerous problem. The leadership strengths that really are at the center of this book are all, we can say, ladder up to vision, but the ability to influence without authority, ethical discernment, exception path thinking, empathy, the courage to kind of challenge stale processes. These are skills that you tend to build when you've seen the downside of systems that are unfair. The winners and the losers are going to be the ones that take this holistic approach. I jokingly called it the calculator, but it's almost like the company operating system in a way. It really is. It really is. And that's what I was reading the tenets of the book and things that you've been saying. It really becomes IT can't run your laptop for you. They put the software on it. They wrote the operating system. If a company is going to succeed, then like you said, these department heads have to own it. They've got to operate the system and be accountable to it. It's absolutely an operating system. Model access is not going to be much of a moat. Everyone's going to have powerful models. And there's so many powerful models to choose from. And there's so many vertical AI solutions that solve very specific problems that all of us could use in different companies. Your advantage is going to come from what sits around the model. your proprietary context, your trusted customer relationships, your understanding of your unique workflows, your better feedback loops, your faster learning, your judgment to know what should remain human, your ability to understand the context, the social relationships of your customers, of your stakeholders, your ability to think of second and third order consequences that the machine doesn't think of, your ability to sniff out bias that the machine isn't recognizing is inherent in its data and training. You're right, that's all the operating system. And that is where the advantage is going to come from is the leaders and the companies who recognize that stuff. It's called The Bias Advantage. Talk to us where everybody can find the book and find more information, Liat. The book is available on Amazon. And if you'd like to learn more about it, you can check out my website, liatbenzer.com, L-I-A-T-B-E-N-Z-U-R.com slash The Bias Advantage. A lot of value here in 30 minutes. I appreciate it, Liat. It's been a pleasure talking with you. Thanks for having me, Ryan. Hey guys, you want to find us, ryanisright.com. You'll find all of the highlight clips. There'll be a lot from this short episode, but let me tell you, it's full of them. A lot of leadership tips. This extends past AI. There's a lot of things here that you can learn to think about it across the landscape of your business, large or small. We've got to start thinking different. It starts with reading the bias advantage. We appreciate Liat. We appreciate you. We'll see you next time or right about now. Here's the truth. Information doesn't change your life. Execution does. So don't just listen to this episode and move on. Take the idea. Make the call. Launch the thing. Fix the problem. Build what you keep talking about building. For more, follow Ryan Alford on Instagram, at Ryan Alford. And watch or listen to every episode at RyanIsRight.com. This is Right About Now. Now quit waiting. Go win.