TRANSCRIPTION 🔗
speaker-0: Welcome back to Fast Break, where we bring you actionable insights at the intersection of business, leadership, and innovation. I'm your host, Matt Cranny, and today we're diving into a topic that's everywhere, but maybe still not fully understood in practice. AI in business. Let's be honest, at this point, most leaders have moved past what is AI. The real question now is, what do I actually do with it? And what does it mean to the future? ⁓ for myself and my business. And so to help us break that down, I've got a fantastic guest today. Doug Barton is the director of the UW EBC at the University of Wisconsin-Madison. Doug, welcome to Fast Break.
speaker-1: Hey Matt, great to be with you.
speaker-0: You can't look anywhere else right now without seeing AI on the front page of ⁓ online publications, print publications everywhere. But to help us set the context, can you share with our audience about your background and and the incredible work that's happening at at the UW EBC?
speaker-1: Yeah, happy to do that. And you're so right. This moment feels extra special. It feels like the world has taken notice of these two letters AI. But just to lay the foundation, my background's in ⁓ accounting and finance. That's what I studied. That was my profession. I would help CFOs and those in finance departments plan for the future, to conduct processes like performance management, reporting, and planning and forecasting. I I would later then pivot in my career and be of service in the areas of optimization, data science, and artificial intelligence. And those included days at IBM and IBM Watson. So I've seen a few of these, what we might call technology adoption and hype cycles before, you know, kind of from the inside out. Yeah. ⁓ before landing here at UW Madison in twenty twenty. I'd love to say a word about our organization, what I get to do here on campus. ⁓ so the UW EVC was born in nineteen ninety-eight. ⁓ it was when the deans kind of across UW Madison came together to form this consortia dedicated to learning about the future of technology and its application in Wisconsin industry. ⁓ really an important time if you can imagine the internet, what it would mean for e-commerce, e-business, etc. ⁓ but that was 28 years ago. And today we're running the same playbook really for digital innovation, digital transformation, and artificial intelligence. And but you know, look, this is really overall in the spirit of service of our UW Madison as the flagship university to impact the livelihoods and the lives of Wisconsin citizens. You know, we call that the Wisconsin idea. And so we've probably been of service for nearly three decades.
speaker-0: That's amazing. That's amazing. And and Doug, one thing that you had shared with me before we started to record today was sort of, you know, your your take on ⁓ the perspective that your role at the UW EBC gives you. Can you talk a little bit about about that too? Yeah.
speaker-1: Yeah, I I really do think that that's what's ⁓ why I'm excited to be here. Cause I get to share from that unique vantage point, ⁓ with engagement ⁓ with eighty more than eighty member companies. ⁓ and let me just kinda unpack that a little bit. So we have a member based organization, ⁓ but we really call ourselves a peer learning consortium. You know, the notion of peer learning, you you can kind of think of that there's curricular or sometimes known as kind of book learning, where you know, you have a particular domain that you're gonna unpack systematically, you're gonna learn from an expert, you're gonna then ⁓ try to apply that work. Peer learning is more about role modeling and possibility modeling. It's it's been said that we often see what's important to us and next for us through the sharing of other people. Who are like us in similar circumstances. So imagine 80 member companies, ⁓ like ⁓ American Family, Northwestern Mutual, Coles, Harley Davidson, Sargento, many Wisconsin manufacturers are included here as well in those 80. But more than 400, 4,500 people participate in our communities of practice. So we get together in kind of professional. Titles. So imagine marketing, customer service, supply chain, technology, HR. That's kind of the model. Communities of practice gathering more than 70 events a year, 4500 people. ⁓ and it kind of comes back to again, you don't even know what's worth learning about until you've seen it through the sharing of another person. Yeah.
speaker-0: one of the things that we talk about at M3 a lot is your proximity determines your perspective. And I I think hopefully our audience and listening to you ⁓ talk about your organization ⁓ gets them excited because this is your your perspective grants you ⁓ this opportunity to have a really unique perspective. And and that's kind of what I want to dive into actually with our first question, if that's okay. And yeah obviously you're you're working, you know, with a myriad of different industries and companies. ⁓ but when it comes to AI adoption ⁓ if we sort of ⁓ I I loved your phrase used before like a hype cycle and maybe in the past. Where are businesses really at right now? ⁓ and you think about this moment in time.
speaker-1: Yeah, it I would definitely say they're further along with ⁓ tools and maybe access to tools than they are with the transformation that I think is ⁓ ultimately implied in this technology. ⁓ you know, the word transformation can be a ⁓ a bit ⁓ lofty, ⁓ but it but it simply means that the way you're it it changing the way you're working in the future. So nearly every company we work with has copilots, right? And chat bots in people's hands. Yeah. Very few have re in a remarkable way changed how the work itself flows ⁓ through their enterprise. One of the things I I I think it's worth sharing about, you know, ⁓ I I I also watch what what's happening more generally. It's been reported that you know, well over forty percent. I've seen some surveys as high as sixty
Load full transcript
percent of US workers say they use AI at work, but ⁓ but the ⁓ officially sanctioned numbers of corporate adoption often tops out way well below that. Mm-hmm. And and this is something to be vigilant for. True. You know, we saw this with social media, we saw this with mobile devices. You know, people were bringing their own device to work and doing work on their devices. That's gonna be true for AI too. When the rewards of using AI can make me more productive and I might be able to reclaim some part of the unpaid overtime, that's not an official statement by me that I work unpaid overtime. I I just want to say that it is often true, right? So so when a technology can have that amount of impact, it tends to drive adoption, even though it might not be sanctioned and governed by companies. So it pays to be vigilant for that and to work with. the technology and your people. ⁓ so to make sure that those gains aren't aren't ⁓ I'll call it captured privately, kept secret, one worker at a time, right? Because we don't want to create a stigma or fear around disclosing we're using technology to accomplish our work. Instead, we want something different. We actually want to have it be scaffolding, ⁓ reaching for a higher ceiling, bringing up the floor for all our workers. ⁓ and we only do that by in some respects working together ⁓ to to accomplish that. You you
speaker-0: Call-out there between ⁓ you know, tools and transformation, I think is really a key takeaway. And then, and then second, sort of the idea that AI transformation is not only happening in the context of the four walls of a business. It's happening to us as society, right? Because AI is everywhere with the service providers we choose to work with. ⁓ even, you know, with for those of us who have children in education and ⁓ thinking about schoolwork and You know, what does an essay mean anymore and all all of those things. And so I love your call to as a business just acknowledge that people are likely all over that spectrum and come into that with all kinds of different thoughts and feelings and responses. And one of the things that because of all that, ⁓ we've definitely seen leaders, you know, of businesses maybe get overwhelmed because AI feels like it can do everything. ⁓ but maybe not knowing where to start becomes the the barrier. And to your point, maybe the start has been everybody has Copilot, everybody has Claude, everybody has Gemini, everybody has Chat G. But let me ask you this. W where do you tell companies who are seeking transformation not to start?
speaker-1: Yeah, I I will I want to reframe this a tiny bit. You know, this this is not my job to tell companies that authoritatively from my perspective, but I will share plenty of times. We've heard from member companies that share privately with their peers kind of what lessons they've learned. So they're the th think of these as maybe anti-patterns, right? Things not to do, right? That ⁓ for success. But yeah, you know. It's been said you don't start with the technology, start with the problem. This is as true here as anywhere. ⁓ there's so many risks to to actually thinking that you have a solution to a problem that hasn't been well defined or defined by your people, by by the amount of pain, suffering, volume, right, variety that that and challenge it creates. One of the things I heard recently from one of our members was. their own reflections on they took us back to twenty twenty three, which was just after that Chat GPT moment in November of 2022. They they and they actually had a a ⁓ banned list, the don't do list. Sure. I'm gonna suggest don't do that today. Yes. You know, today we should have a few clear ground rules. ⁓ you know, you think about protection of intellectual property, ⁓ ethics, you know, those are important ground rules. But when you put out a a list that that is longer than any human would read about things not to do, you're sending a clear signal that you you don't want adoption. So guardrails over, you know, the idea of a ban list. ⁓ I also wanna say, I mean, this may be obvious, but you don't start at the riskiest part of your business model. You know, don't start where, as some say, where errors are expensive. You know, keep your pilots away from the high stakes. ⁓ we've had some of our members keep them away from, let's say, directly customer facing roles. Yeah. ⁓ so a lot of the earliest work were I'm gonna use a large language model or a generally a general pre-trained transformer. That's what GPT stands for. And I'm going to ground it with data that resides in my customer service. ⁓ database or in my standard operating procedures that I know is trusted already and therefore will provide you know essentially the grounding so we'll eliminate or reduce hallucination from the the generally pre trained transformers. So that was a way, frankly, to get started to start to compound the results to learn how we're going to bring the the this technology into ⁓ use cases for customer service, proactive customer care as examples. ⁓ but they were serving our internal employees first. You know, don't start with the risky, but start with the safe and also start where there's a lot of volume, where there's clear feedback where you can start to, you know, experiment ⁓ with with less exposure. I also just want to say one point about this. There there is like these you know, I sometimes c refer to them as Goldilocks conditions. So Goldilocks famously chose the porridge that was just right temperature, right? You know, while you want to avoid the risky, you don't want it to be trivial. Right. Trivial projects are gonna be the death of your important initiatives. It's just signals that this is a toy, this isn't for prime time, and that's not true either.
speaker-0: No, that that that's really amazing, Doug. And and I kinda wanna use one of those into a launch point of a different question. You know, our CEO here at M3, Mike Vickerson, has talked about, you
know, the impact of AI, and we've said it before, is likely the biggest change management exercise he's had a chance to lead through. ⁓ certainly it's true for for me. You saw the ⁓ the rash of CEO transitions that were announced ⁓ at some major organizations. And one of the things that was talked about there was, you know, Is that somebody saying, hey, I'm not the right person to lead us through this next chapter of AI? And so whether that's true or not, certainly I think we could acknowledge a a a a huge change management exercise. And so can you speak to some of the the best practices that you know you're seeing with organizations who are getting that part
speaker-1: Yeah. Right. Yeah, it it's so there's a number of conditions you need to get right here. ⁓ one of the largest companies in the world, Walmart, had a CEO succession, and it was really important as the incoming new CEO just declared that AI will not ⁓ cause a reduction in our workforce. He was sending a clear signal that the people we have are the people that will lead us through. Yeah. Right. That we will not take the gains from artificial intelligence throughout this massive, massive company and turn it into a reason for layoffs and job loss. What a signal that sends. What a trust that builds. All of a sudden you have partners, right? Instead of merely employees that worry for their employment. So trust is a precondition to all this. Technology is the easy part of the spend. It's the parts that go around the technology that create the some regard it as a nine to one ⁓ index, meaning for every dollar you spend on tech, imagine nine going to change and in this case probably process change. ⁓ because there are things that we should just never do again because of the drudgery, right? Or the opportunity to automate. ⁓ but let me get back to the things that I see in common. ⁓ I I see visible personal executive sponsorship right in the companies that are doing it well. ⁓ explicit commitment, productivity gains won't be converted into layoffs. That's just a lesson from Walmart and so many others that are that are longing to build trust and then to build with their employees. Have named champions ⁓ in every function. You know, if you're, you know, typically organized, ⁓ you might have functional leads. ⁓ you might have business unit leads and then also create peer-to-peer sharing, almost like an employee resource group emerges on how we're gonna manage the individuals, the teams, and the functional ⁓ transformation. And do think about readiness before rollout, right? And that that is an important ⁓ part.
speaker-0: Yeah.
speaker-1: I did want to relate. ⁓ we had the good fortune of having a gentleman, Ethan Malik, at our annual conference. He wrote a great book, ⁓ New York Times bestseller called Cointelligence. He's got a mnemonic that I I like or a model. He called it leadership lab and crowd. ⁓ and what he meant is leadership has to kind of define where where we're going, right? They paint the destination, they also set the incentives. And incentives are important here, right? He he observed that. A lot of use cases are going to bubble up from people really close to the work. And and this is really important because not all AI is equally good ⁓ at jobs or tasks are which are parts of jobs. And then finally, you you almost need a lab to harvest the the good insights and innovations and scale them. Right. It won't be one individual that scales, but it's kind of a lab who's going to harden. the use case and application and then extend it. So I just want to offer that leadership lab and crowd has been a something I return to frequently ⁓ in conversation ⁓ to be helpful as well.
speaker-0: That's a great recommendation. Doug, could you rec could you say the title of the book and the author again so our audience could grab that? Yeah. I know I'm gonna be ordering a copy after our conversation today. Yeah.
speaker-1: Yeah, it's a terrific ⁓ book. It's called Co Intelligence and the subtitle is Living and Working with AI. And the author is Ethan Mollock.
speaker-0: Awesome. Thank you. Thank you. Thank you. I I love that framework ⁓ Doug. And and you know, one of the things that ⁓ we would and you referenced it before, as we talked about the the incredible possibility of this moment, but also the the challenges of this moment, right? And we've talked about the change management aspect, but I also want to talk about sort of the risk of AI, whether it's, you know, data, whether it's accuracy or or maybe just back to our organizational resistance. What concerns are you seeing from your, you know, members and and and executives that you're engaged with on sort of that topic of just addressing the risk?
speaker-1: the the biggest thing is there there's a worry that AI is a data or intellectual property security risk. Like how do we control you know that data security? You know, many of these tools are cloud based. Some of them now are actually open source and can be installed actually in your data center, should you want to do that as well. So there are options that are emerging for your technical teams. But I think that's what leads most is that there is a worry by using these tools. You're somehow giving up the crown jewels of ⁓ your enterprise. ⁓ I have to say for the most part that those are are not ⁓ well founded ⁓ challenges. They're overcomable ⁓ by the licensing constraints, by working with your cloud vendors. I do think that the the worry about accuracy, ⁓ And the workforce anxiety that that that comes along with it, what are we going to do with the productivity gains that we get? Those are the the ones that you have to design for. You know, the way you design ⁓ out the the the worry and risk of inaccuracy is you design processes to keep humans in the loop and you keep them attentive. People will begin to trust it perhaps too soon. ⁓ and because the performance is jagged, there are some things it never gets wrong and other things
it will get wrong ⁓ you know, ⁓ on a random chance. Just imagine how hard that is as a human to have a ⁓ you know, something they have to watch for, right? ⁓ rigorously, but they were they they may not have the you know what what I'd care characterize is the you know the vigilance. They catch it when it's wrong. ⁓ people will stop checking. So that that is a worry. Yeah. Those are kind of legit legitimate risks, but but they can be addressed through good design of process change and also ⁓ system.
speaker-0: Yeah, no, that that that's perfect. Doug, what let's talk about the other side of the coin if that's okay. So sure. ⁓ I I try to be ⁓ as optimistic as I can and a realistic optimist is ⁓ is something that I aspire to. ⁓ you seem the same way as it relates to AI. Can you talk about your optimism about what this transformation can do for our businesses, for our communities, the news is sensationalized, it seems so much. Absolutely. ⁓ and that this, you know, a lot of people that can trade on fear. ⁓ and so maybe just take a a couple of minutes to talk about your optimism for this moment.
speaker-1: ⁓ absolutely. I I ⁓ you know, it's told best. I I'll I'll I'll recount, for example, a contribution by the chief technology officer of Brunswick. But the chief technology officer, Anye Denari, ⁓ spent 45 minutes on stage at a recent event recounting the many ways they're using artificial intelligence in their products and in their processes. So if you can imagine this. Part of what they want to do is engineer their marine products to bring more joy to people and to eliminate the challenges they have. And so she was able to share with us ⁓ autonomous docking for boats as an example. I don't know if you ever tried to dock a boat in a windy condition. It just can tear your ego apart. Yeah. Yeah. In in safety, ⁓ mind you. But it's a it was a stunning example about the About the human ingenuity, thinking about the application of artificial intelligence to a better product experience. What's more is then she could also talk about the effect on supply chain, how they can reduce their inventory of parts and yet make sure they can deliver to ⁓ their customers, or maybe managing a challenging trade environment with tariffs. ⁓ so these are all examples she could talk about on how artificial intelligence has greatly enhanced their products and also their their processes. That's what we hear more than anything really across our community is that we're putting it to use to reach a higher bar in what we provide our customers, whether it's a service ⁓ or it's a product. And That will be good for our competitiveness. What's more, we we also know that our processes sometimes can be clunky. There's a famous author that had this quip when asked about artificial intelligence, he says, I'm all for it, ⁓ because it's the bulwark against natural stupidity. ⁓ he was talking about humans, yeah. Right. That we do make some silly mistakes and especially when we're working together across extended supply chains, etc. There's just is so much to be gained. by the increased level of intelligence and coordination that can eliminate waste. I'm really optimistic that this will make our ⁓ productivity higher, it'll make our organizations, our enterprises more resilient and competitive. And that'll mean more things for the people that work in those companies.
speaker-0: Yeah, I I I love that. You know, one of the things that we've talked about at M3 is, you know, our our bet, similar to ⁓ Walmart and and so many others, is that, you know, AI is redemptive, not replacing. And so how can we continue to lean into, you know, there are some elements of our business that are very manual in nature. You know, we move data around a lot ⁓ from our clients sending us data in various formats to getting it into our system to getting it into our insurance companies. If there is AI that can help us. With that, that then it'll frees up our people's time to pick up the phone, call their clients proactively and say, Hey, we haven't talked in a while. Is everything okay? How are you? That's sort of our bet. You know, do we think if we can become more effective, that people's jobs and experience here get better? We do. Do we think it improves the experience that our customers have with us too? We do. And so these stories of encouragement and optimism are definitely out there. And I'd encourage all everybody who's listening today to to really think about what Doug just shared ⁓ as it relates to their business in terms of how can you leverage that for the betterment of all of your constituents. Doug, if I can, I'd really love to, you know, if you are bumping into your one of your, you know, 80 member organization CEOs comes up to you in the corridor and says, Hey, Doug, this whole AI thing, ⁓ I maybe I've given the the tools, but I'm really working on the transformation of all the things that you've shared, and there's been so many great. pieces today, Doug. So thank you. What would be the one that you would say to that CEO, this is what you need to focus on in the next ninety days?
speaker-1: It it might depend on where they themselves are. You know, for those that haven't experienced AI, ⁓ they deserve to buy themselves a twenty dollar a month subscription to one of these tools and bring it to work with them. ⁓ because it's really hard to lead when you don't understand. Right. ⁓ so the the way to learn the frontier is to test it on your own work, give it a whole problem, you know, not a keyword. You know, this isn't a Google search bar. Yeah. You know, the these tools are actually there to understand context, reason with you, to use tools on your behalf to accomplish a goal. ⁓ so for for those that haven't started, I just say get started and bring AI to everything you do for a month. You'll be eminently better informed. For those that are a little further on and might be thinking about organizations, I would encourage them to do something,
you know, first Think about some core workflows that have enough of pain and therefore opportunity for improvement that they can bring a team to, a cross-functional team. ⁓ remembering that, you know, AI won't automate the entire thing, but there are particular tasks that probably act as bottlenecks, right? Yeah. here's the other thing ⁓ for those that are a little further on, is if you've if you've not asked who's using AI. ⁓ and people haven't volunteered it, please declare amnesty and offer recognition and incentives to have people come forward with their use cases. There's a reason they were holding that back because they didn't know what you would do with the information of them using AI in in their daily work. There are some other best practices that just go along with technology adoption. I'm really proud that when we often have member companies share their frequently thinking like this is. Measure the baseline, pilot, measure the improvement, ⁓ take stock of what you learned along the way and iterate. So keep improving and keep improving your improving, right? This is how you stack and compound the games. Yes. ⁓ that's that's really important. And I think you know, not not everyone's been trained in that sort of
speaker-0: Yeah.
speaker-1: ⁓ I would have called it ⁓ in my IBM days kind of agile or lean startup ways, but it is the way you're you're really just compounding learning. ⁓ and everyone has to be real good about sharing the lessons that come out of it and I I think, you know, humble, ⁓ yeah, proud and humble at the same time.
speaker-0: Doug, that's what a perfect way to end this sort of part of our conversation. I I think ⁓ you know, those are ⁓ really wise words. And I hope for everybody listening today that you're hitting pause. You're certainly not listening to this ⁓ conversation at 1.5 or 2.0 speed. This is a 1.0 ⁓ conversation Doug is sharing so much that's so relevant and so helpful. Doug, what I'd love to do is just Actually, can we give you some rapid fire questions for us to finish with if that's okay? Absolutely. And then and then we'll ⁓ we'll wrap. I'm curious, ⁓ and I'll preface this one. This is not so quick. ⁓ when we looked at what our most common ⁓ ask of our ⁓ co-pilot was across our organization, it was organize my day. Help me organize my day. ⁓ and that speaks directly to your piece about that can be done on a Saturday ⁓ for personal lives, right? That can be done in in work. And so with with that as the kind of setup. What is your the biggest AI productivity tip you share the most, whether it's a personal one or a work one?
speaker-1: Yeah, I I have to say it may not be productivity. It's productivity in a maybe a grander scale. I I've taken to with some of the most effortful work I engage in, I I ask the AI to take the perspective of who's gonna get this message. Like take that perspective and tell me how I can improve this. That's so good. ⁓ in some ways it it it it's like having a partner and editor with me. I I get to be more empathetic to the receiver. And that is something that, you know, frankly, I don't know. I I'm I'm at that point in my life where I get a little ingrained in my ways. Sure. And this is a way of holding the mirror up to, you know, the quality of my thinking, the clarity of my communication. And it's a gift to the people around me. Yeah. Not a bad idea. No. Right. ⁓ rather than kinda thinking about my productivity, but thinking of everyone's productivity in a new way. I love
speaker-0: Bet I love asking it to take the perspective of your employees, your clients, your and and giving you those what a what a really valuable insight. ⁓ really quick, the biggest mistake that you see companies making with AI right now.
speaker-1: Yeah, yeah, I think this is a temporary one, but but thinking about tools versus the ⁓ vers maybe tools only versus tools and the redesign of work. ⁓ so I if I have a worry, it's not just tools, it's tools and the redesign of work, ⁓ and the whole value creation engine of our company.
speaker-0: I I love it. I love it. One skill that you think will matter most in the AI enabled world.
speaker-1: ⁓ you know, it's ⁓ I wouldn't be the first person to say this, but management. So defining outcomes, verifying outputs, being clear, that is what management's about when you're directing people, right, to their highest calling too. And so management is kind of that superpower in dealing with a more agentic era. That's
speaker-0: Super, super helpful. And maybe last one, ⁓ from me, one word that describes the future of AI.
speaker-1: I just come up with bright. It's bright.
speaker-0: I love that. What a what a great way to end our conversation. Doug, this was a a true pleasure. Thank you so much for making time to spend with me and our audience today discussing this. ⁓ as we said at the start, ⁓ this incredibly important moment ⁓ at a really important time. And I think you've been able to shed some really clear light and maybe just illuminate the path a couple of steps ahead for people listening today. ⁓ and so I am deeply grateful to you. So thank you so much.
speaker-1: Hey man, I want to thank you for the work you're doing too. ⁓ you know, I I we have this phrase on our campus, together forward, forward being our state motto, but together being a particular way to go forward, it's a critical time to lean on each other. And I hope I've been helpful. And ⁓ yeah, best of luck to all your listeners.
speaker-0: So for everyone listening, if there's one takeaway from me, ⁓ and I'd be curious to hear from all of you in the audience what your big takeaway, but for me, it's not about the tools, it's about the transformation. And using that transformation to create real business impact for the people that look to us ⁓ in all the ways that they do. So thank you for joining us on Fast Break. If you found this helpful, be sure to share it with a colleague and we'll see you next time. This has been Fast Break, brought to you by M3 Elevate. Do you want more tips to help grow and protect your business? Subscribe now and catch all of our episodes. I'm Matt Cranny. Thank you for joining me, and we'll see you next time.