From Hype To Habit: Making AI Work At Work

Back by popular demand, this encore episode breaks down how to turn AI hype into real results.

Aritificial intelligence shouldn’t add more work—it should make work smarter. Jon Duren and Jay Custard share how to turn AI hype into real results: integrate it into daily workflows, tie pilots to KPIs, manage adoption like a product and design for trust. A playbook for making AI actually work at work.

Support for this episode provided by: Digital Realty

Learn more about this week's guest:

Jay Custard has deep experience helping companies harness the power of digital technologies to drive meaningful customer experiences. An experienced digital marketing and ecommerce executive, he has worked across multiple disciplines with both large and small companies to enable integrated transformational retail experiences on a global scale. He has been fortunate to serve the customers of Crocs, New Balance, FinishLine, Southeastern Grocers, and Cabela's with a focus on user experience, analytics, marketing, merchandising, and technology transformations.

Jay's top pick: 4 Key Principles for Reimagining Retail with Science and Vision

Jon Duren is an accomplished technology professional with over 25 years of experience in service-provider, data center, cloud and AI solutions. He is driven by a passion for applying technical solutions to achieve business outcomes and is currently working to expand the WWT AI Solutions practice. Jon actively seeks out the best strategies to create customer value. He is constantly monitoring marketing trends, innovative and disruptive technologies, and significant market transitions to identify opportunities for driving customer growth and success. With a talent for breaking down complex topics into easy-to-understand concepts, Jon is able to develop and share a vision of where business and technology intersect to create better outcomes.

Jon's top pick: Agentic AI Sounds New—But Haven't We Been Doing This Already?

The AI Proving Ground Podcast leverages the deep AI technical and business expertise from within World Wide Technology's one-of-a-kind AI Proving Ground, which provides unrivaled access to the world's leading AI technologies. This unique lab environment accelerates your ability to learn about, test, train and implement AI solutions.

Learn more about WWT's AI Proving Ground.

The AI Proving Ground is a composable lab environment that features the latest high-performance infrastructure and reference architectures from the world's leading AI companies, such as NVIDIA, Cisco, Dell, F5, AMD, Intel and others.

Developed within our Advanced Technology Center (ATC), this one-of-a-kind lab environment empowers IT teams to evaluate and test AI infrastructure, software and solutions for efficacy, scalability and flexibility — all under one roof. The AI Proving Ground provides visibility into data flows across the entire development pipeline, enabling more informed decision-making while safeguarding production environments.

Chapters

• 0:00 The Real State of Enterprise AI
• 1:12 Why AI Suddenly Feels Everywhere
• 3:20 The Weirdest AI Use Cases Yet
• 4:40 How Consumer Habits Shape Workplace AI
• 7:50 AI Hype, Fear and the Hard Truth
• 10:16 The Real Reason AI Adoption Stalls
• 12:20 Off-The-Shelf Wins and the 25% Rule
• 15:20 Stop Chasing Shiny Objects—Start With KPIs
• 18:20 Why IT Should Lead—With Security First
• 20:00 Shadow AI: Threat or Secret Weapon?
• 23:00 Too Many Tools, Too Little Impact
• 26:20 ROI Reality Check: What Actually Matters
• 30:00 Unlocking Hidden Value With Product Thinking
• 33:20 Why Trust (and Friction) Fuel Adoption
• 36:20 The Playbook: Fast Pilots, Real Results

Transcript

SPEAKER_00: 0:00

The adoption of AI and generative AI tools within organizations is a complex and evolving challenge. As AI tools and features become ubiquitous across your business, there are a few key considerations you should keep in mind to drive successful implementation and capture value. That's why I'm excited to have John Duren and Jay Custard on the AI Proving Ground podcast today. John, an AI and data solutions expert, and Jay, a chief digital advisor, tackle this challenge from different points of view. But one thing they agree on: by taking a thoughtful, human-centric approach, organizations can unlock the transformative potential of AI while mitigating risks and driving sustainable value. AI is a journey, not a destination. And as you'll hear from John and Jay, those who navigate it most effectively will be poised to thrive in the years ahead. This is the AI Proving Ground podcast from Worldwide Technology, everything AI, all in one place. I'm your host, Brian Felt. Let's get started. So a couple days ago, I was walking through the San Jose Airport. I noticed an ad from our friends over at uh Dell, and it said, AI, putting the AI in IPA, the beer, the beer IPA. Which got me thinking, so where's the weirdest place you've seen AI show up so far in this weird world of ours?

SPEAKER_01: 1:34

You got me an IPA. Now I'm wondering like, what does that look like? These little bubbles I've been drinking, are they somehow tailored to my behaviors? That's that's pretty fascinating. I think the beauty of AI sometimes is that we don't realize it's all around us, right? I think some of the more shocking places that you find it is when it it kind of jumps out at you from, you know, why are you trying to inject this into this particular experience, right? I know you were trying to eliminate friction, but but now you've created something where I'm now aware of it. You know, I don't know that it fits well in. I was going through an online uh enrollment for, you know, a medical thing for my daughter to get a physical. And then somewhere along the way, you could tell that someone, a marketer, had gotten into this thing thinking, like, hey, I can learn more about these people, aside from the information I'm giving you, and stuck some crazy interstitial in the middle of it that was trying to take me off to a almost like a survey, but it wasn't like you would normally get on a website. It was much more of a different way. But I could tell the intent because we get to live and breathe in this stuff, but that wasn't the place that it was supposed to be. So I find like things like that, where people are trying to gather more information, understand me more as a customer, like those things that are noticeable and pronounced, they do not do anything to help the adoption of AI in my mind, right? So, like I don't need that in a medical enrollment piece to take me off to get more information about you know who I am, what I care about, what brands I want to delve into. So it was just a strange introduction of something that normally wouldn't be there. I'm all for AI and IPA to make that better, get it in distilleries so we get better allocation, but you know, keep it out of my workflows when I'm just trying to get things done. So, I mean that that was an example that I came across lately. I don't know what John may have seen because you know he's behind the scenes in a lot of things and knows more about this than I do.

SPEAKER_02: 3:17

Yeah, I think interesting. You know, you you bring up, you know, alcohol. My first thought was you know, AI in the wineries, um, you know, using it to you know improve the performance of the of the grapevines and so forth. But yeah, I think the most interesting one I've seen was probably one of the cities in Texas is putting a digital human on their city services page. Interesting. That's that's a really interesting effect. And I'm not yet sure that we are ready for digital humans.

SPEAKER_01: 3:47

So we'll see. I'm old enough to remember, John. Do you remember the days back in maybe the mid-2000s where there would be a random person that would walk out from the bottom of a browser, like when you were trying to look for like blinds, right? I'm reminded of that. It was this kind of disassociative random ghost that would pop up, and it was just this, you know, a properly dressed person in a business suit trying to, hey, can I help you along? It's like this is the weird manifestation of Clippy in my WordPress or my Word documents, right? Like, I'm with you. I think that jarringness of trying to introduce digital humans or digital like humans when they're still kind of 4K and glossy. I don't know where humans are ready to integrate that into their day-to-day, right?

SPEAKER_00: 4:27

Yeah. Well, Jay, well, I I love the idea of you shopping online for blinds. I love that that was your um example that you gave.

SPEAKER_01: 4:34

Where do people shop, man? I don't I don't go to a store to look for blinds. Come on.

SPEAKER_00: 4:39

John, what does that mean? You know, what does that say about the world that we live in today from maybe from an enterprise perspective where AI is becoming as much of a feature, as much of a product within things that we already own and use on a day-to-day basis, as opposed to just something that, you know, maybe a year or two ago we thought about we have to build all this unique stuff. Um, what does that say about the state of enterprise AI that it's popping up everywhere?

SPEAKER_02: 5:04

I think it's actually really beneficial. I mean, the the biggest challenge I consistently hear from customers is not about the actual technology, it's the adoption, getting people to use it. Okay, if you're gonna spend this much money building out an AI tool, an AI infrastructure, how do I ensure I get the people you know in my organization to take full advantage of it? And I think the answer to your question is the more we see it in consumer goods in our everyday lives, the more comfortable the average employee is going to be with you know checking into the technology at work and using it and playing with it to try to learn how they can can improve their lives. But I I do believe that that's it, that's key. It's actually going to be key to all of our customers is in seeing that the employees are using this outside of work for things that can improve their lives themselves, dramatically improves their their likelihood of using it at work.

SPEAKER_01: 5:59

Yeah. Yeah, it's kind of funny how Gen AI became shorthand for AI and all of a sudden it's like, oh, it's everywhere, and now nobody wants to know what to do with it. It's threatening my job, and I don't know where it is. When in fact, we've been living and breathing and relying on AI for such a very long time. But when it showed up in this weird package that did these things that it had never done before, and it was kind of a shortcut to this, and now there's this world of, you know, you know, either super optimistic and things are gonna be, you know, we're not gonna have to work in three to five years, or you know, it's gonna take all of our jobs. Like there's there's this dissonance. But man, how long have we relied on Google Maps and Waze and Siri and all these automation things that have happened behind the scenes and dealing with Spotify versus Netflix versus whatever, showing you the next show that you're gonna sit there and watch on a Friday night and not get up to go get your AI-inspired IPA? Like, I think now that we're seeing this kind of become more prevalent, it gets that adoption curve within enterprises to kind of take away some of that weirdness or some of that unknown or unknowable. And we just get more familiar with it. And I think that's how enterprise grade tools and the way companies start to inject this into their business flows and their day-to-day, it becomes that much more natural to them. Right now, with the introduction of any new technology, it's just jarring and it's weird and we don't understand it. And it becomes kind of this thing in and of itself, and it really hinders adoption, except for those early adopters. I'm old enough to remember when I got my first email address when I was in college. I was a junior in college, and it was not better than me being able to interface with my professor directly. It was not better for me to go get my grades from the registrar or pay my bill. But I think every technology that gets introduced gets applied to current methodologies, and it's usually not as good as what we've all been used to. But now we're starting to see these things roll into usefulness and adoption and ways that people are starting to kind of bring these things and get comfortable with them. And it's just going to be a rocket right from here, I think. But it's still scary, right? And it's still weird.

SPEAKER_02: 7:59

Jay, let me ask. I mean, do you believe, really believe that AI is going to be more transformative than some of the technologies you've listed off?

SPEAKER_01: 8:09

Oh, John, you're putting me on the spot. So I no, I don't. I think we're in this hype cycle of all kinds of things that could be, right? And I think this one is such a jarring extension of a lot of the things that we've come to rely on and things that we've gotten used to using. Connectivity, the ability of our phones to keep us connected socially or too social for my teenagers who don't really interact with human beings, right? I think it's easy to look at this technology and the way that it kind of jumped onto the scene and kind of surprised everybody and the adoption of how quickly this kind of came because the adjacent possible was there. We had the infrastructure, right? Think about the introduction of the internet. It took a long time because we didn't have fiber, right? We didn't have all the connectivity, we didn't have data centers, we didn't have this network of connection around the world. This is just a screaming example of something pretty revolutionary that's been building for a while. We just didn't see it. And now it's here. Uh, I don't know that I'm gonna go cash out my 401k and pretend that I don't have to worry about money because life is gonna be easy for me in four or five years. But I also don't think that the world's gonna end in that same amount of time either, right? So this one just is jarring because it's so easy and it's conversational. I think the use case and the way people interact with it, right, John? I mean, when Gen AI is weird because you can get basically a glorified Google search back really, really quickly, and then I can start to apply it to things that maybe I don't want to do in my job every day. So you're kind of like, wow, potential impossibility is here.

SPEAKER_00: 9:39

But well, the adoption one, you know, it's interesting because on a global scale, Gen AI has been, you know, almost you it feels universally adopted from a consumer standpoint more than anything we've seen so far. But you know, ironically, you know, at least as we hear from our clients, you know, we we surveyed uh a group of IT leaders recently, and half of them said their biggest challenge was uh adoption of new technologies such as AI within their workforce. John, why are we why are where's the disconnect there? Why is is such a struggle for companies and their employees to adopt this technology?

SPEAKER_02: 10:16

Well, it yeah, that's an interesting approach. So I I think one of the things we're seeing more from a trend perspective is that AI has been used to date to really create yet another new tool that folks have to use. And that's that's an uncomfortable place to be in. Okay, now I have yet another place I have to go to do something throughout my day. I I think we'll see the adoption improve as features and functions begin to be built into the applications that companies and employees are using throughout their normal business day every day, instead of it being a separate tool. Today they're not built into the applications you use, they're a separate web interface in many cases, and someone has to remember to go there and they have to have a reason to use it. So adoption is a little bit more difficult. It's it's something extra for me to do. When that chatbot is just built into the application that I use every day, which you know, whatever industry I'm in, I have a list of applications that are core to my business at that point. Then it becomes just part of doing my job. And so the adoption will go up. We'll also see a lot more return on investment, a lot more value out of the tools when it becomes part of the integrated into the daily daily processes.

SPEAKER_01: 11:34

John, do you think that's because companies aren't looking for a problem to go solve with this new tool they have, but they're they're figuring out how this tool actually fits into existing challenges with the business? I mean, where is that shift really kind of come in? Is it just familiarity with the tools, or is it uh, you know, starting with the solution and working backwards, or is it our company starting to work from the problem to the solution that just happens to be a chatbot, for example?

SPEAKER_02: 11:59

Yeah, I I think the answer to that is the level of maturity of the technology, of the companies using it, et cetera. You know, I I don't want to look down on chatbots. I still think they're fantastic. They are an ends to a means for a lot of things, and they are, for most of our customers, a learning experience. It's a great way to dip your toe in the AI pool and start learning what generative AI can do, is capable of, and so forth. But until we get to the point where we really take the time to start integrating into other other tools, other applications, it's it's a learning experience. And the interesting part there is it was required. We had to go through that learning experience. No one, you know, two years ago, generative AI wasn't a thing. And none of none of our customers, no employers out there just automatically woke up and had tremendous amounts of generative AI skills on stack. And you know, so we're we're all going through the learning phase together.

SPEAKER_01: 12:52

Yeah.

SPEAKER_00: 13:22

Do you think that some of these um, you know, off-the-shelf or you know, AI solutions integrated into Teams or uh Kodium or, you know, AI into any solution that you might use from the enterprise, is that is that a little bit of an easy button to start to creep towards that 25%?

SPEAKER_02: 13:40

Absolutely. I'm a matter of fact, I think that's exactly where we're headed, especially over the next 12 to 18 months, is is a lot of off-the-shelf work, a lot of just turning on, evaluating and turning on the features in those core applications is is exactly how we're going to see companies begin to reach that. I think there is a caveat in that study that we need to remember, though, is you know, turning on new tools is great, but the real power of AI, when you hit that, say roughly 25% and the the value stream starts really turning, you know, turning up in a big way, that's when you're really beginning to transform processes, the way people work, not just the tools that they use, but actually transform the way people work. So the tools are going to make incremental steps toward that 25%, but in and of themselves, they're not fundamentally transforming work. I think some of that transformative effort has to happen in conjunction with you know buying off the shelf. Jay, I mean, I suspect you see a lot of development in tools as well. So yeah. For sure.

SPEAKER_01: 14:45

No, I think you're spot on. I think the really interesting part is, you know, the unique and the thing that I love about worldwide and working with John and all the folks on this is that we know that that 25% is super valuable, but we also know the totality of just finding a way to get a toehold to even get to that 25%, right? There's a so much struggle behind that for companies to figure out a mix of shelfware versus you know things that they're going to build on their own or they're going to borrow from others, right? There's so much churn underneath that. And I think the thing that, you know, John sees it from the hardware side and the and the way that these tools are being developed with, you know, native AI tooling in them, right? Can't ignore it in Microsoft products or service. Now it's fine, right? It's coming. That helps the adoption curve. But there's also this underlining piece of companies are starting to figure out the right problems to solve with this technology, which gets you faster to that 25%. And that is like the proving ground, right? And then by then, you've got the transformation of workforce to your point is now we have an infrastructure. Now we have some talent, now we have some things where we feel secure about these types of things. We can fundamentally interrogate the capabilities of these tools. At the same time, these companies have done a great job of interrogating the implications of these tools on the workforce and on the day-to-day behavior of their employees. And I think it's that value construct coming together that really drives these things into this optimized world where companies can transform their workforce, look for new business models, things of that nature. So it's fun to me to see technology and business being forced to come closer together in this space because that's how you get to that 25%. And that's how those companies are going to become the standard in their industry, and they're going to leave everybody else in the dust until someone can figure out how to do that unlocked to John's point.

SPEAKER_00: 16:35

Yeah, Jay, I want you to uh unpack a little bit more the idea about getting that toehold. Um because easier said than done to get to that 25% mark where you start to see the tipping point. What do you know, how would you advise um clients or organizations? How can they start to go about um considering, you know, these tools that have AI functionality within them in order to drive up that adoption? Like, what do they need to consider? How do they need to be approaching it? Um, do they need to just like go out and ask their employees, like, hey, what are you using? Maybe we'll start from there. What's the process?

SPEAKER_01: 17:07

That's a great question. Um, I think what we've seen is an evolution of that, right? If you want to like beat the old Henry Ford thing to death, right? You go ask your employees what they want, they're gonna tell you they want a faster horse. And like, you know, again, like there's some truth in that. I think what we've seen is when generative AI in particular came on the scene, right? It took all that nerdy stuff that data science to, you know, could do, all the algorithms, the heavy math, all the power of AI that we've been using forever. It exposed it to all of us and it changed the paradigm. But we went looking for problems to solve with that rather than staying tight to the strategic initiatives and the roadmaps that were already established within an organization. So John and I have seen this where companies go out and they create individual KPIs to prove the efficacy of AI. When in fact, why wouldn't you start with some of the business measurements that you're already aligning your company strategies to, your hierarchy, your hiring, your talent, your investments, your return on all that stuff is already in service of your business model. How can you take those, deconstruct those into the core components and look for ways for any sort of technology, if it happens to be Gen AI, great. How do you start to layer that back in so it shows building drivers of those major macro KPIs that most organizations are tracking living and have to report to a board or shareholders or whatever that looks like? So, you know, I think it's a long-winded way to say everybody looked at this tooling and technology as very being very deterministic. It was going to give me the answers that I needed. So we looked for problems that maybe fit that. And what we realized with Gen AI is very probabilistic, right? It doesn't give you specific answers to things. So I think understanding just that capability and then applying that to the right parts of the business allowed us to figure out how something like a chat bot or an RFP assistant or some of those things kind of come to life. But they weren't metrics or opportunities or strategies built on their own. They were a component of the macro strategies that companies were already trying to drive for, which makes it that much easier to find inefficiencies in the process or new opportunities for workflows. And then I know that feeds right into how John and team help size these things and make them then fundamentally come to life. I mean, am I out of out of line, John? What am I missing? No, you're right up.

SPEAKER_02: 19:23

But I I do want to jump on something you said just before that little part. You actually said something very, very uh I think we should should cover the intersection of the value of AI with the value to the business. I think an interesting thing we've also seen, and and part of what our teams really spend a lot of time with customers on is in traditional technology, IT has been able to connect those dots.

SPEAKER_01: 19:48

Yeah.

SPEAKER_02: 19:49

When you start talking about IT uh AI, it involves a lot of aspects of the business that that the IT organization in many cases has a lot less visibility into. So connecting the value stream to the value of the technology is not necessarily as easy. And we're finding that it results in effects of things like buying product, hoping to find a use for it, or just buying general high-level overviews and looking for use cases, not really understanding them. I think that's one of the interesting challenges about this technology is it's created an easy paradigm for the IT departments because now they've got to figure out how to really understand parts of business that they've never had to look at before. I could buy an Ethernet switch, I can buy a router, I can buy a server, and I know what impact it's going to have on the business. IT or AI is a lot less intuitive on that, just at face value.

SPEAKER_01: 20:45

Yeah, I agree. I think there's this old story, right? I kind of think in stories and parables and things like that. There's there's a famous one that there was a building in New York, right? High rise, very well, you know, all the great occupancy rates, all that stuff. They got constant complaints from the tenants that the elevator was too slow. So, what'd they do? They went out, right? They were afraid of losing tenancy, they were afraid of, you know, having you know, like most of our commercial real estate, they were worried about losing that rent. So they went out and put it done to some engineers to figure out what we can do to increase the capacity of the elevators, right? So they went through all the things. Can we do bigger cars? Can we do different algorithms that make cars come in here? Can we put things through? So they accelerated from an engineering technology problem. How can we make the elevators faster? Well, what came back was a study that said, hey, this is gonna cost an exorbitant amount of money, right? You're gonna have to invest, it's gonna tear this thing down, you're gonna have all this disruption, you're gonna have to do these things. And the landowner said, Well, we can't do that, it's not feasible. So they gave it to an intern who happened to be there, who was a psychology student, who was trying to get his way into the commercial real estate space. And you know what they did? They put$200 worth of mirrors in the lobby. Because the problem wasn't that the elevators were too slow, it was that people were bored and couldn't have anything to do to distract them while they were standing there feeling the pain of this. I think there's analogy here around we can go do all these things and put this on our IT teams to go and tell us how to size for all of this stuff. But if it's the wrong problem to solve, it's never gonna find traction inside an organization. So it's a silly example, but I think it's really the differentiator between the companies who have adopted these technologies the right way because they're asking the right questions, and then they can right size the solutions to John's point to solve against those and give them the momentum by which, you know, to continue to grow the business, but through the right lens, right?

SPEAKER_00: 22:38

So, Jay, is that where dynamic persona modeling and value stream mapping comes in, really kind of laying the groundwork and a path towards sensical ROI and not just uh, you know, ready aim fire?

SPEAKER_01: 22:52

Yeah, I think that's a great place to start, right? So I think anything within an organization that has kind of been a heuristic by which we've tried to solve challenges or you know, we've gone to the measure of call it what you want, right? Persona building, customer journey mapping, and employee mapping, things of that nature. I mean, that is fundamentally where it starts. The cool thing is that AI can actually assist with that, right? So you get past kind of the uh the explicit of what people tell you, but can also evaluate in a massive amount of information, data, utilization of tools, you know, broken processes. Like our manufacturing friends have done this for years, right? Like Six Sigma lean methodologies kind of have this built in. If we can start to apply that into kind of this human side of things within organizations, not only looking for process breakdowns, but looking for maybe talent breakdowns or you know, the inefficiencies of people with a mental bias that I don't want to automate this because that means I'm not relevant anymore. Like trying to find those things is where companies have to go, at least in some level. Being vaguely right is better than being precisely wrong in the situation. And I know John has seen this and counseled a lot of our customers, and I know he has a lot more to add around how people are actually doing these types of mapping and how this stuff starts to come to life into the practical application of you know a roadmap, right? That gets you to these tools. So, you know, John, tell them tell them how it actually works, man. As if I had the magic bullet. I know, right? We'd be on our own doing something else, wouldn't we?

SPEAKER_00: 24:24

Yeah, John, well uh you know, maybe weigh in with what role does it have in that in that process? Are are they now in a position to lead it more? Um, just given the stakes.

SPEAKER_02: 24:34

I I think this is a chance for IT to show more relevance to the organization if they're involved, but leading it's a tough, a tough call. Facilitating, probably a better word for for IT. That way they have they have the ability to weigh in on the decisions, they have an understanding of where and where and how the technology is being used. But in most of the the transformative type use cases we've seen, the business, the line of business really has to lead the charge. They have to have a need and they have to have a desire to solve the problem. And then IT needs to be brought along. And I really want to, we cannot leave out the security parts of IT as well. Probably the single most thing we the most important thing we've even neglected to say here, but security absolutely has to be a part as well. But I IT, the benefit is going to be in this world that IT gets an opportunity to truly connect the work that they do every day fundamentally to a specific business outcome, to actually map it exactly to changing and transforming the business. And that's something I think that's that IT has struggled with for my entire career. You know, IT technology for the most part is the plumbing and the electrical wiring in the house. It's great. You have to have it, but no one buys those things because they want to launch a whole new business or a whole new service product line. AI actually can do those things. It can actually be a part, directly a part of launching entirely new products for an organization. In the BioLive Sciences, for instance, using AI to identify new drugs, literally to create new product for the organization. Things that IT never could do before, now it has a hand in. So I think that's the role we're going to play.

unknown: 26:26

Yeah.

SPEAKER_00: 26:27

Happy you brought up cyber, and I had a note to bring it up uh later in this conversation, but I'll go ahead and ask this now. If AI is everywhere and it's in all the things that we're doing, isn't that just inviting shadow AI into our organizations, which is going to create massive uh cyber concerns?

SPEAKER_02: 26:45

Shadow AI is already here. It's not invited. You know, it's uh I think Chat GPT probably was the single biggest, you know, uh shadow IT-related technology we've we've seen in my career, um, just because of the rapid, rapid adoption of that technology. But yes, we're gonna see some of that. But I would say at this stage, I and I I want to be very careful, not saying everything shouldn't be secured. Absolutely. IT and the security should have a hand in everything. It's this shadow, shadow AI that's probably at this point helping us drive the adoption problem, helping us get past the hurdle of the fear, the uncertainty and doubt associated with AI, and become more comfortable so that they will be more likely to use the solutions. We're also at a very, I hate the word immature, a very early stage with AI in most enterprises. So the security and cyber teams are playing catch up. They're trying to get to that point so that shadow AI will be less of a problem. But it's you know, it's it's a problem across all technologies we have. This is nothing new. It's just a little exacerbated right now because this is so new and so immature in the space.

SPEAKER_01: 28:00

John, you could not be more correct. It is here and it has been here for a while. And people like me, who used to run e-commerce businesses or an omnichannel business on a retail brand, we built it and were incentivized to build it as fast as we could. So, I mean, the reality is you're right, this is peeling it back. We've all had, you know, personalization engines and loyalty engines and email engines and all these things and all these businesses. Take it outside of retail. This stuff has existed inside domains across industry for a very, very long time. But there are business impacts of that in the sense that, you know, it's net integrated. We would we would do everything in our power to stay away from an IT prioritization process or going through a formal piece. I could go buy a point solution that was that much faster, that would get me good enough information to get something done. I think now is the time where AI has created such a noise around just visibility of it, that companies are forced to kind of look at this ice this estate holistically and realize the dependencies of these tools and of these vendors they bring on all these other parts and pieces to where this comes out. I think to John's point, shadow I shadow AI used to be shadow IT, but it's really shadow AI, really has helped kind of create the business value to go in and show a CIO or a CEO that, you know what, we've been doing machine learning and predictive analytics and a lot of AI stuff for the last eight years. That's why this part of our business has grown. Like there's a comfort with that. It's a heuristic that people can put their mind around. And then it becomes a better conversation between IT leaders, operations leaders, risk leaders, financial leaders, business leaders to come together and look at these things holistically, not only to drive more value out of the information and data that these tools provide, but to secure it so that we don't lose something in a blink that took us years and years and years to build, right? So I think John is spot on. It's it's unnecessary we talk about it as like it's an evil, but it's actually going to speed up adoption because it's already in parts of the business. It's now a time for IT to become much more strategic along with their cyber friends. And not think of these things as bolt-on or just the folks who move information around, but those that are super strategic who actually help develop new ways of growing and retaining customers regardless of industry. I think it's a really cool inflection point. It's going to be painful, but we see it every day where things are changing, right? I mean, John and I spent a lot more time together. I don't know any the first thing about HPA, but I've learned a lot from John and I'm trying to bring that back in around how businesses actually make decisions. That's what we do at worldwide, is try to decouple these things and make them work together in a much more harmonious way. But a lot of lot to do, but I think super exciting and a forcing function that probably we all needed, quite frankly.

unknown: 30:40

Yeah.

SPEAKER_00: 30:41

Well, to borrow another term from cyber, from IT in general, um, you know, if we're bringing in these concepts of more AI, are we at risk of just over-tooling ourselves again? I mean, that feels like a uh a lesson learned over the course of time over and over again.

SPEAKER_02: 30:57

I and the risk is always there. And yes, organizations are going to over-tool, probably overbuy, probably enable more features than they need. But we'll we'll see that you know that there'll be a uh you know a leveling point where they begin to work on correcting that and focusing on where the real value is at. You just have to remember right now we're the technology is still young. I mean, generative AI, while AI in general has been around for years and years, generative AI really is what brought brought this out, made it viable for the the average enterprise to focus on it, to make it part of the business priorities. That's less than two years ago.

unknown: 31:39

Yeah.

SPEAKER_02: 31:39

And you know, the technology has to continue to mature rapidly. In that time, we're gonna learn a lot, we're going to over, we're gonna do some things wrong, we're gonna do some things right, we're going to overpurchase and we're going to underdeliver. But at the end of the day, the companies are learning, they're getting better, and the technology is going to prove to be both a business enabler.

SPEAKER_01: 32:00

I agree. I think it's kind of game theory, right? That there is there's a there's a cost to be in this game, but you have to be in this game right now. Something of this magnitude is sitting on the sidelines and waiting is absolutely, I think, the the only losing strategy. Now, do you have to go headlong into it and you know do the massive build-outs of certain infrastructure? I mean, that's that's up to an individual business in terms of how they ally their risk. But, you know, understanding these tools, really thinking through customer workflows, employee expectations, things of that nature are to John's point really, really important. And we are going to get most of this wrong. Like that happens when you have these big step change moments in technology. It happened during the internet boom. It happened during the mobility boom when everything had to be mobile first back in the day, right? So I don't think we're ever going to escape that. Um, the best way to predict the future is to kind of make it. So I think companies who are aligned to a strategic vision of who they want to be in a few years have to rationalize how these tools fit in and make some bets. And some of them are not going to pay off. But, you know, understanding that downside risk versus not participating at all, I think is a really key piece. And I think that's something that at worldwide, it's why we've invested in the things that we have, to de-risk some of these things, either in proving grounds and whatnot, or having people like John to come on site and you know bring information and tear these things apart, but always through the lens of what is it meant to do? If this were an employee, what would you hire them to do? And if you didn't answer that question, probably don't need to be investing in that, right? So, you know, I know that's the approach these guys take. And I think that's why we're starting to see some of these companies turn the corner on how to get in the game, but not risk everything on the game, right?

SPEAKER_00: 33:39

Well, Jay, you mentioned, you know, companies should try to identify what they're going to be or what they want to be in a few years. We can't even, you know, anticipate what's happening in the next couple of days or weeks. How can a company be flexible and you know, how can they, you know, be firm in what they want to accomplish, but also be flexible and adaptable for whatever is coming down the line? I mean, that feels like competing forces.

SPEAKER_01: 34:03

Yeah, is this the place where I put my QR code up so someone can sign up for my own consulting business where I could tell you all the magic?

SPEAKER_00: 34:10

Blinds. I think we we mentioned blinds is a good business for you. Right?

SPEAKER_01: 34:14

Yeah. Um that's it's a fantastic question. And I guess where I become more of a realist around this, I'm still AI optimist in terms of what I think it can do for the world. But at the same time, I think companies exist. They have inertia, right? They have momentum, they have an install of a customer base, they understand their employees, they have probably a good market position now. I think starting and reverting back to who you are and where you were planning on going is the first logical step to kind of circle the wagons. Like, let's make sure we're all aligned on the things that we have put to our shareholders. Every company I know has a three to five-year plan. If you have a five-year plan or beyond that, I think you're, you know, maybe a little bit in trouble. Three-year plan, right? What are those major movers that are going to impact your business? Where do you think that the puck is going? As you've been doing it for years and years and years. And then how do you use AI tools? One, to critically inform that, right? You can find new research, new understanding, right? Really stress test some of those potential risk models, investment models, things of that nature through whatever means you want to. But then it's also like, how do you put this technology in service of that? I know there's a lot of companies who are very progressive who are already figuring out new business models, right? John brings up, you know, life sciences. There are drugs that have been discovered completely by AI that are now going into clinical trial for the first time. That's amazing. But those code folks have been there, right? So I think for most businesses, it's getting back to what's important, what are the things we think move the needle, and then trying to find ways by which any technology can help do that in a way that moves them forward, rather than trying to just say, we're gonna be an AI first company, we're gonna go shove this down, or we're gonna completely change our business models. That's a pretty good risk, right? And you're gonna need some help doing that. But I think just fundamentally understanding the three or four things that move the needle for you and how these can come in service of it is the only way to move forward. It's got to be small change. Shoot off some bottle rockets, right? Don't launch a space shuttle. And I think that's what we're seeing a lot of companies do right now.

SPEAKER_00: 36:15

Yeah, John, you know, maybe bringing it back to the near term or at least the you know, the present day. Um, let's say a company has normalized on the amount of tools that they're using as it relates to AI. Um, what about integration? That seems to be something that's also popped up from clients is in terms of concerns is you know, how do I get these various turn uh tools uh to work together? So, what are you seeing from an integration standpoint?

SPEAKER_02: 36:41

That's actually easier than than maybe you would think. I mean, over the last I don't know, eight to ten years, mobile app development has really driven the world toward APIs. So almost everything we see now is built with a strong foundation in APIs that makes integration easier. And if you look at a lot of the AI off-the-shelf type products and the ones that come in the big packages, like like Copilot from Microsoft, uh Einstein from ServiceNow or from Salesforce, things like that, these are have well-developed APIs specifically designed for those integration pieces. And I think we're gonna we're gonna see a lot of that. And it's not gonna be a massive uplift from what they're doing. The harder parts are getting the features implemented and getting the adoption. The integration pieces, I don't think, are gonna be major hurdles for most organizations as they get going. Now, for those who have no developers on staff, uh, that might be a different story. But AI is going to require some new skills that maybe they haven't had before, and this could be a good area to invest.

SPEAKER_00: 37:47

Yeah. What about budgets? Um, you know, if you're continually adding more and more tools, I mean, certainly realizing that, you know, IT spending as it relates to AI is is going to be going up, but you know, everybody's looking to save. What is this going to do from a budgetary perspective?

SPEAKER_02: 38:06

Well, tools rationalization and and oh overall is an ongoing problem for I think every single customer we meet with, you know, especially in the technology space. So that's gonna be an interesting, interesting to see how this plays out. I don't know, we have a great answer today. We're still it's still so early in the game. Um, I I do believe we're going to see AI tools particularly, but AI displace a lot of the tools we've had before. I mean, I've even seen and heard of of solutions out there that are AI tools that are looking to displace big, big packages like SAP. You know, and if if if an AI tool can do something like that, we know it can displace vast numbers of tools across the organization. So I think we're going to see probably in the near future, not this year maybe, but within the next couple of years, uh, an effort to focus on tools rationalization, you know, evaluating where we can consolidate tools, where we can eliminate tools, how AI is changing the game. And it's it's going to be just an ongoing process, but it's not a new thing. It's something we've been through in cycles in in IT over many, many years.

unknown: 39:17

Yeah.

SPEAKER_01: 39:18

Yeah, I think you're right. I think even just the forcing function of having this be such a prominent board-level initiative and everybody trying to get into the game. At some point, you kind of have to look around the estate and you have to kind of you know go looking in the corners, like, what do we actually own? And what are these things actually doing? And how do you bring this conversation across your operations and your business teams and everybody else who's really kind of running these parts and pieces? It's it's a new time, right? To get alignment among all the right leaders, not just an IT exercise to go tell me what I've got from an asset perspective, but it's an actual like, why do we have these, right? You really kind of dig in. I think that is something the companies are doing. They're starting to discover where they may have some bloat. But to John's point, I mean, this is as old as IT as it's been, right? It happened during the internet boom, happened again during mobility, right? It's happened along the way. Um, but I think these are moments where it does make sense to bring teams together to really understand the value of what they have, jettison what you don't need, and be able to invest in the things that might set you up for a potential win going forward.

SPEAKER_00: 40:21

Yeah. Well, most of what we've been talking about so far seems to be um leading itself to somewhat of an ROI conversation. So, you know, maybe maybe just bluntly, how can organizations um articulate potential ROI related to these tools? And then once they have them implement it, what are the measures of success or KPIs that they need to be looking for to help prove that out?

SPEAKER_02: 40:47

You know, that's probably if you just read the news on AI, that's probably the single biggest question the media is constantly asking right now. And the answer is it's different for every user. And I I can there's some commonalities that I can share. I would say up front, it's probably more important to spend time early in the process of use case identification, understanding what your company, maybe your rubric is, how are you going to evaluate the priorities of these things? It's easy to look at something of you know cost versus effort, but when we start talking about AI, there's so many more factors involved in the actual return on investment. Uh even for instance, just how hard is it going to be to you know to get the adoption necessary, just which we've already talked about. So I think spending time up front and the development of the use case, I would expect you know, rock solid companies have processes for building business justifications for new processes, for changing processes. This would follow any of those typical processes. There's nothing revolutionary new about the way you approach an ROI, but putting the rigor behind it and forcing it to happen. This is not just buying a new brand of server. This is a technology that's going to transform something. So it needs that business justification behind it up front. Probably the biggest mistake we would see made today is companies who are jumping to the end of the story and trying to buy hardware to do something rather than spending the time up front to focus on that use case prioritization and the business justification. But those who do are today seeing the return on investment.

SPEAKER_01: 42:30

Yeah, John, I couldn't agree more. And I think it's there's there's almost like this reckoning that we've seen customers go through, right? That there's there's two kind of paths generally, right? Is are we trying to completely innovate or are we trying to optimize, right? If you want to go down the innovation path, then the only thing that you can do to kill that is to say prove it, right? Because if something is truly an innovation, it cannot be analytically proven before it exists, right? So, and that's people are gonna take a leap of faith, right? No one could, you know, decide on the the innovation that was the iPhone. Like, is this going to make money and be the most transcendent consumer electronics device in history? No CFO that said prove it. That thing was never gonna go forward. Most companies don't have that appetite, but I think we're kind of at this inflection point. If you're truly looking for real innovation, new business models, things of that nature, you're gonna have to do some things differently outside the prove it model. I think where most companies find themselves, to John's point, is how do we optimize? That's you can apply some risk analysis or some financial rigor to that. I mean, we have existing business case type logic. I think where it needs to start is a collaboration between business unit, employees, operations leaders, whatever that might look like. Not an IT exercise to go justify this, but what is this going to do? Like find those metrics that we all know move our business. Look for those in decomp decomposed ways by which what are the drivers of my revenue goals or my GP goals? How do you continually break that down and find inefficiencies in that and make some bets and model out with some sensitivity analysis around we think an improvement here in this and you know a 30 or 20% range gets us to this much value, which then leads to this, right? I think there's ways to do it, but what most companies get stung stuck up on right now is they're trying to innovate, but they want to prove it before they innovate on it. Or to John's point, they just jump ahead and then try to back into it. And we don't see those folks being successful. The ones who are successful, decompose the metrics that they're already watching, measuring, and using that are part and parcel of how people show up and get paid, get their bonuses, or get fired, right? Take those metrics, decompose those and figure out holistically how the entirety of the organization can support that either through the lens of generative AI, machine learning, or just simply a sticky note somewhere, right? It doesn't have to be complicated.

SPEAKER_00: 45:05

From what you're seeing, I guess, you know, John or Jay, how much underutilization is out there right now in terms of you know AI capabilities in software suites that you know we could be leveraging but just aren't or don't know about?

SPEAKER_02: 45:19

Oh goodness. Um, I mean, even just looking internally to ourselves here at Worldwide, there are so many features that I know our our employee base isn't even aware of today. Uh, you know, different different parts of the organization, everything from marketing, trying to build AI tools into the way we build and add articles and things and images to the platform, to technical writers and the way they work on the backside to help us with RFP tools, RFPs and responses, document summarization, et cetera. Um, today the utilization is a small percentage of where it really, really needs to be. And I think every organization is going to see that. I I back to the point though, the easiest adoption, I think, are going to be when features are added to the existing packages. You know, I and I like healthcare. I'm I'm have a familiarity with a lot of the healthcare. You know, you're looking at McKesson tools or Epic or Cerner, you know, some of these big packages that healthcare organizations use. The users are in those packages all day, every day. Turning on features just makes their day job easier. Adding a tool like a chat bot on the side is something extra. And I know that's repetitory, but it's it's interesting from an adoption standpoint to see these features being added. How do they evaluate them? We've talked about when do they evaluate them? I think the time is now. Jay had a fantastic point. I think probably the most salient point of this entire session. The game is being played and they have to be in the game. You know, there's going to be an expense, there's going to be a cost, we're in the learning phase, but not playing the game is at a much greater risk than and than anything you may lose from the change in the time, the experience and education you're gaining by being there.

SPEAKER_01: 47:07

Yeah. I think what gets lost a lot, and I'm a I'm going to tell you transparently, I am one of the biggest non-users of some really powerful tools that we have in our organization right now. I'm not going to name them, but I don't use them because they don't fit into my daily workflow. Right. I haven't figured out how to change the way I do my business to accommodate or find value in that tool because there was, you know, kind of an onboarding piece and it was, it was opened up. That's not going to stop, right? AI is a feature in every one of the tools that comprise the enterprise employee suite, right? It's there, it's in service now, it's in Teams, it's everywhere, and it's not going to go away. And to John's point, there's good in that because we're going to get used to it and we're going to start to figure out more about what works and how it works. But I think there's a miss because the reporting that I've seen, I know we've looked at these things internally, is that it's usually a who, what, when, where, right type of thing. We can, we know that really, really well. What we don't get to is why. And I think there is this missing component as we roll out these tools, and I'm focused solely on employees and turning these features on, is that there's almost needs to be kind of a product owner mindset or uh, you know, a design-centered thinking that has to go into that. Like why? Why are we doing this and what is the benefit? There's a lot of features, but there's often not a lot of benefit that's conveyed to that end user around how to do it. So if you think about the great product companies that are in the world, the Netflix or the Spotifies, like they have teams who do nothing but turn on features all day long, but they're always doing it through the guise of what is the benefit to the end user. I think some of that thinking needs to kind of be brought into bear in terms of how companies think about where these tools add value, how they add value, and then quick quickly pivot off of that. I think Teams is going to be a fantastic tool with Copilot down the road. I just haven't figured out how it adds value to me because I often don't need some of the things that come out of that. Or if I don't have all the connective tissue for certain parts of the thing, because I can't get to this repository of information yet, then I don't need it. And but I don't want that to be taken as a sign of this tool isn't good. It's just, it's not there yet. It's not mature yet. It will be better. And I think we can make some really quick judgments on these and pull things back, but I don't think that's the full picture. We need more product and design-centered thinking because you got to put humans back at the center of this, even though AI says take the humans out of everything. Well, no, right? There's still people at the end of the day using these tools. We haven't turned into ones and zeros yet, as I can tell. Maybe John, because he's so damn smart. But you know, there's other things that we're going to have to figure out from an empathy and a humanity perspective of how behaviors evolve as we introduce these capabilities, which I don't think we've ever seen in the history of humankind. It's going to be fascinating. Well said.

SPEAKER_00: 49:51

Yeah, Jay, I want to um dive into that a little bit more because I love the fact that you just brought it back um to a human-centric mindset. Um and earlier you mentioned how AI can be jarring. And you know, when something is jarring, you're turned off, you don't use it. So if we're looking to grow adoption to get to that tipping point, but if you unleash just a wave of new AI tools and it's jarring and it would turn off some people, like what what do companies need to do to make that more of a smooth landing so that employees are ready to use those tools if and when they're given them?

SPEAKER_01: 50:28

Yeah. First, hire John, right? Bring him on board for all these consultant engagements. It'll help me get this sorted out. Uh, in lieu of that, um there's an underlying thing. I'm gonna take I'm gonna take this from a different direction, right? I'm fascinated with this concept that in my mind, AI, in particular gen AI, always equates to speed, right? The underlying assumption when people deploy these technologies is that I'm gonna move from here to here faster, and I'm gonna skip over a bunch of steps because that's gonna be commoditized, right? So I'm taking friction out of the system. I fundamentally believe that people need friction at certain points to help them understand that I have moved from this to this, or I feel good about moving from this to this, right? The analogy that I would give you is I used to run e-commerce for a massive outdoor retailer. And what did we do? We tried to figure out how to take steps out of the checkout process all the time. It's all we spend our time on. Let's get to one click checkout, let's get to this, let's get to this. And we did it. And guess what? Our conversion rate went down because we made people mad. Because they wanted a chance to review, they wanted a chance to feel like they were part of this. I think a lot of these tools that we're introducing in AI right now are under the auspices, directly or indirectly, efficiency, speed, get to this faster, right? Do more. And I think that fundamentally makes it hard for humans to interact with these things because I got a bunch of information that kind of got regenerated back to me. It didn't come from John in accounting or, you know, Mary and this part of the business who I can trust. Now I gotta go through this and I gotta review it. Like, is it hallucinating? Is it weird? Like, why is it here? How does it know this? So I think there's an element of that where AI can't always mean speed, it has to mean efficiency. And that sometimes means introducing friction into these things so people feel comfortable in their heuristics of how they adopt and bring these tools to bear in a way that solves a problem rather than wow, we just jumped to the end of the story and I missed the four, you know, the five chapters in here. I didn't really learn anything, but yeah, whatever. Guess what? Now I get more tasks to go into, right? I think there's something human about that, right? I mean, John, am I crazy?

SPEAKER_02: 52:43

No, I matter of fact, I that deserves an applause, man. Well said.

SPEAKER_00: 52:48

Uh well, John and uh Jay, um, you know, thanks for joining. Any closing thoughts or or you know, thinking back on the you know, the the the hour or so that we spoke here. Um, what did we learn? What are we taking away here?

SPEAKER_02: 53:03

Uh I'll start. I'll say, you know, we didn't really cover it, but one thing that was of interest to me, our CTO in a recent presentation said, and I I I repeat this to many of our customers, as you move into your AI journey and we're all on this journey together, design realizing there's going to be mistakes. Gosh, we're so still so early. And that's the biggest, most important thing. Nobody's going to get it right the first time. Matter of fact, in perfect example, even worldwide's own internal chatbot development effort, I would say just over the months that we've been working on it, the underlying technology has already changed like four times. Be prepared to deal with the change that's happening, the evolution of the technology, month to month, almost week to week right now, and understand things are just not going to work out some of the time. And we have to learn from the technology and move forward. But the movement and the journey is so powerful, and the the the results are going to pay off in the long run in big ways.

SPEAKER_01: 54:06

Yeah, I think John, maybe you said this once before, but or maybe I'm summarizing at one point, but be very, very disciplined and deliberate and and not inflexible about the expected outcome, but not on how you get there. Because anytime you introduce a technology like this, it's inherently going to be maybe worse than what it's replacing, right? I can't imagine that when cars kind of rolled out that people were like, yeah, this is way better than a horse when there were no streets, there were no road signs, there were no gas stations, right? There were no wagons. Like, how do I carry stuff? Like I say that because I think we jump ahead to all these final conclusions and we try to read the tea leaves of where we're going, but innovation is absolutely messy. We had this thrust upon us, and we've got to figure it out. So, one, understand the difference between innovation, truly breaking things, because you can't prove it in advance, or being methodical in terms of how you do that and being grounded in the things that are more core to your business to satisfy your customers and your employees, and you can find a path forward. Define those outcomes and be inflexible around what you want to have happen. But man, find a partner, and even not us, anyone where you can test this stuff. So the solution is gonna get wiggly and squiggly before you get there. But own that outcome that you want, and you can find ways to make this stuff kind of bend to your will as opposed to all of us, kind of where we're at is bending to the will of AI and trying to squeeze anything into that, which that's not who's gonna win in the long term, right? So, and I I can't say this enough. Like the things that John's team represents, the thing that worldwide has invested in, it helps take the sting out of that. It helps move you along, it helps you figure these things out. So if it's not with us, find a partner, find a place to be able to do this because experimentation and trying and stumbling is the only way to make this go, but you got to be in the game in some fashion.

SPEAKER_00: 56:02

Yeah. Well, great stuff. Jay, John, thank you so much for taking time out of the you know, busy schedules. Um I'm sure, whether it be client partners or just try trying to stay on top of any and all this AI stuff. Uh so thanks again, and uh, we'll talk to you soon.

SPEAKER_02: 56:16

Thank you. Thank you.

SPEAKER_00: 56:18

Okay, a big thanks to John and Jay for taking the time out of their busy schedules for joining us on today's episode of the AI Proving Ground podcast. A few key lessons learned before we leave. First, look to integrate AI into existing workflows. The focus should be on integrating AI capabilities into the applications and tools employees are already using, rather than introducing standalone AI tools, which will help adoption. Second, always connect to business outcomes. Companies need to clearly map AI initiatives to specific business priorities and KPIs, not just implement AI for technology's sake. And finally, embrace an iterative learning mindset. Given the rapid pace of AI evolution, companies must be prepared to experiment, learn from mistakes, and continuously adapt their AI strategies. If you like this episode of the AI Proving Ground Podcast, please consider leaving us a review, or rating us, or subscribe. And sharing with friends and colleagues is always appreciated. This episode of the AI Proving Ground Podcast was co produced by Nas Baker, Stephanie Hammond, and Mallory Shaffrin. Our audio and video engineer is John Noblok. My name is Brian Felt. We'll see you next time.