Beyond the Hype: What Actually Separates AI Winners from Losers

[00:00:00] Scott Kinka: Technology is changing faster than ever, but the most important conversations aren’t about the tools. They’re about the ideas and the people shaping what comes next. On the Bridgecast, I sit down with the leaders driving change—from AI and cybersecurity to the infrastructure powering modern business. Together, we explore how technology moves business forward: what they’re seeing, what they’re building, and what it means for the rest of us. I’m Scott Kinka, and this is the Bridgecast, presented by Bridgepoint Technologies.

Hi, and welcome to the Bridgecast. Today I am joined by Maribel Lopez. She’s the founder of Lopez Research, a Forbes contributor, author of Right-Time Experiences, and host of the AI with Maribel Lopez podcast—we’re going to talk about that. She’s also a regular source for The Wall Street Journal, Bloomberg, and CNBC. Basically, she’s a super cool expert that you’re going to want to hear from. Today, her work focuses on AI inside enterprise organizations, with particular attention to customer experience, where she says AI success or failure becomes visible the fastest. We’re going to get into that as well. Maribel, welcome to the Bridgecast.

[00:01:19] Maribel Lopez: Excited to be here. Good to see you, Scott.

[00:01:21] Scott Kinka: Yeah, fantastic. And it’s always good to interview a fellow podcast host—you’ll be assessing my work throughout this, obviously. Hopefully I do well. Well, let’s get right into it. I’d like to start with your own journey, just so people can get a little grounded in who you are and why this is going to be such a cool episode. You studied finance and marketing, which is, I’d say, maybe not the traditional path to becoming a leading voice in AI. Help us unfold the journey—how did you get started? How did you land where you are?

[00:01:52] Maribel Lopez: Long time ago, in a land far, far away. I actually started in finance at Motorola, then I went into technology marketing at Motorola. We went through that big mobility wave. I did an industry analyst gig, then went back and did another technology marketing gig, another analyst gig, and then I decided to do my own thing after all that. So it’s been about eighteen years now at Lopez Research, and we’re having a good time helping people go through really big technology changes. We’ve been through mobility, we’ve been through cloud computing, and now we’re going into AI. Great fun times. I’m already—

[00:02:21] Scott Kinka: Off script. I told you I was going to be off script, and now we’re off script by the second question, but I have to do this. You said cloud computing and then AI, and I’m watching it happen all over again. Are we making the cloud mistake, rushing in all over again, or did we learn something?

[00:02:40] Maribel Lopez: We definitely learned a few things. For example, when we started with cloud, the idea was everything was going to go to the cloud—everything. And you fast-forward almost twenty years, and we can see that there’s still a lot of activity in people’s data centers and on-premises. When you think about going into AI, instead of everybody thinking, “Hey, we’re all going to go into the cloud,” they start to think more thoughtfully about where their data is going to live, where they’re going to do things, and how much it’s going to cost. So we learned a lot in the cloud journey that I think we’ve carried forward. But now there’s just a ton of new things we have to worry about.

[00:03:13] Scott Kinka: I love the optimistic answer, and I’m not disagreeing with you—I’d agree. At least we can have those conversations now. I’m curious what your thoughts are on this: I think we’ve got a challenge between the C-suite level and the people in IT who actually have to get the work done. Maybe we earned the right for more pushback, because IT was so powerful and central to businesses during COVID and the work-from-home transformation—it kind of earned a seat next to the CEO. But it felt very much like, on the cloud revolution, it was, “Go get some, I’ve got to tell the board.” And I don’t think that’s necessarily changed, at least from the rest of the C-suite—not the technology part. There’s quite a bit of, “Go get some AI, I don’t want to hear about it.” Are you hearing that too?

[00:04:08] Maribel Lopez: Yeah, I think it was actually worse in AI than it was in cloud computing, truthfully. In cloud computing, there was still a lot of dissent about whether cloud was going to be a thing and whether it could even happen. By the time we got to what you could call the OpenAI era—which started with the foundation models and people being able to use it at home—there was definitely a sense that if you don’t do this as an organization, you’ll be left behind. So there was a tremendous amount of pressure to go out and, as you said, get some AI. We saw lots of board mandates that said you needed to be doing AI, and that’s at the big corporate level. But at the smaller and mid-tier corporations, there was still a lot of what I’d call competitive fear happening—you’d see it in the media: if you don’t do it, you’re going to be behind. So I think the push was faster in some ways, and greater. I mean, we could argue we’ve been talking about AI since the sixties, so we could debate how fast or slow that term moved. But the AI we have today, starting with the foundation models and large language models—that era kicked in much faster than cloud did. It took at least five years for anybody to have the same level of buzz about cloud that they have about AI today, which happened in about a year and a half to two years.

[00:05:27] Scott Kinka: Do you think the major difference is that while end users were using cloud products, it was really an infrastructure change, whereas AI became generally available to the end user directly? So if the company didn’t do it, end users were going to do it on their own, and it just billowed up quickly, and all of a sudden businesses had to react?

[00:05:47] Maribel Lopez: The biggest change AI has had on the technology market is that, for the first time in all the years I’ve been looking at technology, everybody can participate in something that is extremely technical, and they can participate at different levels. You can learn how to use software programs you didn’t know how to use. You can theoretically code things when you’re not a developer. So that kind of empowerment, I think, really changed the conversation around whether we should deploy and how fast we should deploy. And just to step back for a second, more broadly: you go into any organization on the planet, no matter how big or small, and they can’t find their information. They don’t know where it is, they don’t know who has it, they spend hours looking through their laptop trying to find the email or the document somebody sent them. Just the power of being able to say, “Dear AI, go find me the blah, blah, blah” has been revolutionary for organizations. If nothing else, just accessing information is really powerful, and everybody wants that.

[00:06:52] Scott Kinka: Fully. You hit on two topics there—now I’m completely off script, but you hit on two things. One was a super easy end-user use case—I love that one. You also mentioned nobody can find their data. Both are clearly true. There’s an ongoing debate I have sometimes on this podcast and with customers: what comes first, preparedness or literacy? What I mean is: yes, nobody can find their data, so isn’t it incumbent on the business to organize and appropriately permission data before turning it on? And then there’s the other school of thought: we’re never actually going to know what the use cases in the business are until we give people the opportunity to get loose. So let’s just build some basic literacy—I’ll use the uncomfortable example—turn on Copilot. I know the two things don’t necessarily have to be mutually exclusive, but when you meet with a new client and they say, “I don’t know what to do,” is it data preparedness, basic literacy, or some combination of the two? How do you rationalize that?

[00:08:04] Maribel Lopez: I’ll step back to where we were and then look at where we’re going. If you go back, say, not quite a year, you’ll remember there were a lot of studies about how terrible AI was, how all the POCs were failing, and nobody was getting any return on investment. It was very much a “sit on the couch, it’ll be okay” moment for enterprise buyers—don’t worry, we’ll get to the other side of this eventually. But why did that happen? I think that’s the context of the question you’re getting at. The reason it happened is that there were at least four problems we saw pretty consistently in organizations. The first was that their data wasn’t ready. The second was that they didn’t know what problem they were trying to solve—they were really experimenting with what the technology could do, as opposed to how to make something in their business better. The third was that they had new metrics and measurements that went with it. And the fourth was the broader category of security and governance. We could break those out and make it five—there are lots of ways we could frame it—but those were roughly the four problems. Depending on which organization you went into, you might have had all four; usually you had two or three. So what I’d say is that, in a lot of ways, AI has surfaced problems we’ve always had. We’ve always had a data-quality problem. I remember when we were doing BI and analytics, we talked a lot about the data-quality problem. We’re in 2026, and we’re still talking about the data-quality problem—it hasn’t changed. Bad data gives you bad AI decisions, so there’s an imperative to get to cleaner data. But if we look at the history of this, people aren’t going to put on the brakes and wait until their data is all clean, sanitized, and perfect for AI. They’re going to go out, figure some stuff out, and make some mistakes. Frankly, they need to figure out how many data sources they have for something—maybe five—and which one is the truth. That’s where data cleanliness and preparedness start.

[00:10:10] Scott Kinka: I love the four problems—let’s orient ourselves around that. We’ve spent a decent amount of time talking about data. Okay, we get it: if we want to do anything significant, or even just train a model internally, we’ve got to decide what data is accurate, what data isn’t, and what processes aren’t documented today—we’ll get back into that at a later date. So data is one. Your second point—and pardon the way I’m explaining it—is the use-case development problem: what do I want to do with it? I’m curious about that, because one of the things I find when I talk to IT leaders—bringing it back to that C-suite conversation—is the C-level saying “go get some AI,” and the IT person going, “I don’t know—do I buy that by the pound? How awesome, but you’ve asked me to solve infrastructure problems, and the problems live in the business.” It’s like we treat IT as the plumber, and now we’re asking them to be the general contractor, but we haven’t told them what the house looks like. The problems sit inside the line of business. Is that something you’ve encountered—that the IT person is trying to figure out governance, but there isn’t really an infrastructure in the business to surface the line-of-business problems that can even be solved? Are we missing that? Is there a gap?

[00:11:35] Maribel Lopez: There has been a gap there. I do think that, if you look at every other technology transition, we always build a foundation first, and the desire is to just go out and build the foundation for AI—that’s consistent with how we’ve always approached the problem. But, to your point, it comes down to: what tools do you need? Is it a hammer? A nail? A screwdriver? An impact driver? There are a lot of different tools you could use to match specific use cases. So I think what people got wrong wasn’t that they were trying to build the foundation—it’s that they had no sense of how to connect the foundation to the original set of business problems. If we bubble this up for any organization, you might have a customer-experience problem, and that’s the biggest thing you need to fix this year. You might have a product-development problem—that’s your biggest thing. You might have supply-chain constraints—that’s your biggest thing. So when you’re rolling out a technology, the first thing it should help with is whatever those key anchor problems are. Those tend to be fairly meta for an organization—maybe they come down into a specific business unit, or maybe they’re broader than that. But what I think happened is we got really enamored with—

[00:12:54] Scott Kinka: All these models—

[00:12:56] Maribel Lopez: —all these frameworks, all these cloud services—it was just a plethora of choices. It was very much like the IoT era, where we built a platform to connect all these different protocols and technologies, and then your manager walked in and said, “Okay, we’ve got a platform—what do we do?” And you’re like, “We haven’t done anything yet, we just connected everything.” That wasn’t necessary, but it wasn’t sufficient either. So I think as we move into AI, we’ve got a similar problem. It was even worse in the beginning, because people were asking, “Well, what could it do?” So we had all these random science-fair projects with AI—not because we were foolish, but because we were trying to figure out the limits and boundaries of the technology. We spent a lot of money and time and ended up with no value, and then management was upset because they told you to go get some AI, and you did, but you didn’t do anything with it.

[00:13:41] Scott Kinka: I guess that gets to your third point, around measurement. You mentioned ROI earlier, and I scribbled it down to come back to it. I’m assuming that’s what you mean when you say—we’ve talked about four problems: data, use cases, and measurement as number three. And by that, I’m assuming you mean sufficient protocols and data that let you pull out productivity metrics, so you can ultimately tell the person who told you to “go get some” that you got some, and it delivered X or Y. Is that roughly where you’re going with it?

[00:14:14] Maribel Lopez: Yes. And sometimes people get wrapped up in thinking it has to be a hardcore number, and that’s not always the case. If you look at any technology trend, we’ve always said it’s going to make us more productive—email was supposed to make us more productive, and we could argue about that one. I spend a lot of time on email; I’m not sure it’s really productive—inbox zero is a whole other issue we could discuss. Let’s unpack measurement. I think the real issue is: you roll something out, and then somebody comes back and asks you what it did. Sometimes it made something faster. Sometimes it saved you money. Sometimes it stopped calls from coming into your contact center. There are a lot of things that could happen. But if you don’t spend a few moments thinking, “I’m going to roll out this technology, and at the end of it, we’re going to see X,” then when somebody comes back, you can’t tell them where you are on the path to X. And X could be a lot of different things. But if you don’t have an X, there’s going to be a disconnect at the outset—because you and the other person haven’t agreed on the terms of the contract, so to speak.

[00:15:32] Scott Kinka: And on those use cases you mentioned—my belief is that if you’re good enough at Excel, you can make your numbers tell whatever you need them to tell, at the end of the day. But you have clear ones: fewer calls coming in, an improved process, more widgets out the door. I think the area where businesses are making their first or largest investments is probably the hardest one to quantify—we’ve been talking about end users just using Gen AI, having access to it. I don’t know that we have the ability, or even the thought process, to measure personal productivity anymore when it comes to how we run HR. So when I’m meeting with a CIO and they say, “I don’t know how to tell my boss that the thirty-some dollars a head I’m going to spend on [insert Gen AI model here] is actually going to deliver anything right now”—I’ll give an example, maybe this will bring it home. Lines of code: my fifty-year-old developer, who’s been doing this forever, produces six thousand lines of code over a period of time, and the new hire we just brought on delivers the same six thousand lines of code in a day. Are we expecting them to produce twenty-one thousand lines of code, or can they go home? I don’t know how we even measure that right now. Are we even prepared, from an HR perspective, to deal with this kind of productivity gap that I think we’re beginning to lean into?

[00:17:13] Maribel Lopez: The question you’re asking is very nuanced and very important, and I think we don’t know the answer to it yet. But let me unpack a couple of things, because you raised a few concepts we should investigate. The first concept is the use of AI. I could be sitting around using Grok to create cat videos all day—is that really a productive use? So this whole idea of counting how many tokens somebody uses is definitely the wrong metric, and I think we’ve started to see that play out at the larger tech companies, where they’re saying, “They’ve used too much AI—that was really expensive. What did we get out of it?” So yes, you want people to use AI, but you want them to use the right AI at the right cost. There’s always this trade-off in any IT discussion—how much does it cost, how accurate is it, how fast is it—and it’s the same with AI. I think we got a little carried away and said, “Just use it, use it, use it,” and now we’re dialing that back. Okay, that’s one. The second thing you mentioned is increased productivity—everybody writing code is a great example, because we hear this all the time. Sure, you can produce a ton of code, but the real question isn’t whether you can produce code—it’s whether you produced better code. Did you take an average developer and make them extraordinary, or were they still mediocre? Did you produce more documentation that helps people find better ways to code? So this gets back to metrics: you can have a metric, but is it the right metric? Faster and more lines of code is not sufficient—it has to be better code, the right code, more products as a result of that code. Code by itself is not an answer, it’s just a thing. Okay, so there’s that. The third dimension you raised is HR, and there are two sides to that. Tell me if this answer is getting too long, but—

[00:19:15] Scott Kinka: HR—there’s how HR professionals use it,

[00:19:18] Maribel Lopez: —and then there’s how we actually evaluate talent. I do think talent evaluation has to change. There will be a component that just says AI literacy is part of your review—are you AI literate or not? I think that’s really important going forward. Are you using AI in the right ways? That’ll probably become something that’s tracked, so what you’re compensated on may change as a result. How you work with AI agents, which aren’t humans, will be another factor that starts to filter in—we haven’t even gotten to that, and I’m sure it will come up at some point. I think we’ll change how human resources evaluations work. And what you do versus what AI agents do has been a big topic of discussion in 2026 and will continue to be, so we’re still working through that. The other side is how HR uses AI. We’ve had some interesting lawsuits about whether AI was screening effectively, and we’re starting to see better model traction around making AI screening more effective. You can also interview somebody at any time now—we’ve seen very convincing voice-AI talent interviews. I saw a demo in New York at a cloud computing conference last month, and it was amazing. That AI could interview thirty thousand applicants—something your HR team just couldn’t do in a short period of time—and do it any time of day, in multiple languages, for multiple types of jobs. It could interview a chef one day and a software engineer the next. It can help you create job descriptions. So there’s a tremendous opportunity in how AI will change talent acquisition as well.

[00:21:27] Scott Kinka: You are listening to the Bridgecast, proudly sponsored by Bridgepoint Technologies. When it comes to customer experience, Bridgepoint delivers innovative solutions that empower your business to stay ahead. Visit bridgepointtechnologies.com to learn how we can transform your customer experience today.

Super good—we talked quite a bit about ROI and expense, and you mentioned some of those clear use cases. But I think, in general, most businesses we speak with are still in the “what can I do with it” category. They know they have to implement something, they know they’ve got to wrap something around it—we’ll get to the governance piece in a second. Generally speaking, I’m going to make a somewhat controversial statement, and I’m curious about your reaction. In conversations I’ve had, I’ll say: this year, maybe the next eighteen months, assume that any of your AI investments are pure cost—just assume that’s the case for now, because it’s going to take time. I’ll stop there—agree, disagree, comment. Is this like a brand-new AI customer we’re talking about?

[00:22:33] Maribel Lopez: Yeah, okay—if it’s a brand-new AI customer, it probably is cost. I think most people I talk to are a couple of years in now, and I’ve known some people who are a good five years in—back before we had generative AI—and they’re killing it right now because they hit all the arrows early, when it was a lot harder and more expensive in a lot of ways. But if you’re coming in now, it takes a while to figure out what use case really adds value for you. I will say, though, I’ve seen some relatively universal use cases for organizations. Most companies have some kind of IT operations support they could leverage AI for—better IT service desk support, automating network operations. We’ve been talking about that for a while, but it’s really gotten a boost in the past couple of years, so that’s something everybody could participate in. Almost everybody has some kind of customer-experience challenge they could address, too. Certainly in the contact center—I don’t love the idea of “containing” calls, because no customer wants to be contained, but there’s definitely a lot of proactive communication you can do, and a lot of calls you can take off an agent by servicing them through AI. This happened to me recently with Hilton: I left something in a hotel room, went through the whole process with AI, and basically got the item shipped back to me—never spoke to a human. That’s the kind of thing that would have taken a while and probably involved some chitchat otherwise; it was extremely efficient for me and for the company. So customer experience is a second area to look at. If you’re big enough, there’s also the HR/talent one—just screening applicants is very helpful. None of these break the bank, and none require a major infrastructure investment—you can get them from your SaaS providers. You’re dealing with Salesforce, NICE, ServiceNow, or Zendesk—your networking company has built all this into their automation stack as well.

[00:24:59] Scott Kinka: InfoTech Research—and I’ll probably get the specific stat wrong—but in their annual CIO survey, they said, generally speaking, that businesses prefer to implement AI in platforms they already own. I guess that gets partly to what you’re saying.

[00:25:18] Maribel Lopez: There’s a very good reason for that. Everything we discussed at the beginning—where you’re trying to build a platform—requires you to know a lot about the business, what it needs, and requires evaluating a lot of technology. Am I going to use a hyperscaler? Am I going to buy and deploy it myself? How many agents am I going to have, and what will they do, and will they cross applications? It gets unwieldy very fast. So it’s much easier if your CRM provider comes in and says, “Hey, we’ve got AI, we’ve got AI agents, and you just have to roll them out—they’ll handle this specific use case for you.” And you say, “Yes, I’d like that specific use case, let’s start there.” Then you go to management and say, “Within ninety days, we rolled X, Y, Z out, and we’re gaining X,” because the SaaS provider already figured out the metrics for you—you just have to deliver on them.

[00:26:13] Scott Kinka: And in most cases, the agents are interacting with the organization’s own data—we’re not connecting multiple disparate sources together, it’s self-contained. If it’s within Salesforce, for example, you don’t have to worry about connecting data sources or whether the data is accurate. I fully get it.

[00:26:35] Maribel Lopez: That’s point four you’re landing on—security and governance.

[00:26:38] Scott Kinka: Yeah, I didn’t want to miss that—thank you for the segue, I inadvertently set that up myself. So your number four was security and governance, and this is probably its own episode. But I’ll say this: I’m surprised by the number of IT leaders I talk to who don’t even have an extension to their existing AUP around AI usage inside the business. What would you say to that person?

[00:27:10] Maribel Lopez: A lot can go wrong here—that’s the first thing I’d say. This is the classic “what could go wrong” scenario. I’ve learned a couple of things over the past couple of years that have honestly been quite terrifying to me. One is how few people really have a solid governance practice within the organization in general—that’s problem one for a lot of organizations, because you can’t extend something you don’t have. Problem two: assuming you don’t have it and you’re moving into AI, you really need to look at testing, observability, and agent and AI security and governance. What we’ve seen in the past year is that none of that really existed all that well until the end of last year, and this year we’ve seen a lot of the technology vendors provide solutions for it. If you were already in a stack—like we were just talking about, your CRM, contact center, HR, or ERP stack—all of the security, governance, and identity attributes extended as you moved into AI. They didn’t necessarily extend as well into agents, because we call agents “non-human identities.” So you’ve got humans and non-humans, and the non-humans have to be permissioned, because they don’t have the same permissions you and I have. Maybe they only do one thing—maybe they only look up whether an item was shipped; they shouldn’t be able to look up your salary information. We have to permission them the right way, and we’ve started to get into that discussion now. So the first thing I’d say is: if you’re rolling out any AI, you have to be really conscious of what data it accesses, and how that data is used, stored, and permissioned. Is it going up into a large language model where you can’t see it, or is it in an enterprise LLM service—what’s the security around that? We won’t dive too deep into it, but there’s a whole category of security threats around AI models and agents that your security team should be evaluating—how secure these services are. That’s one side. On the other side: do we know what these AI agents did? Can we track it? Can we make sure they did the right thing? If they did the wrong thing, can we roll it back? All of that has to be in place too. Job one is to make sure you do no harm. Job two, you could say, is to try to do something productive and make money—but if you get the “do no harm” part wrong, you’re going to be shut down within months, whether by a data breach, a security or compliance group shutting you down, or legal shutting you down. There are a million ways your AI project can get shut down, so there are certain things you have to take care of right out of the gate.

[00:30:26] Scott Kinka: Is this a good spot to bring up something you said on your LinkedIn profile? We saw this: “The best technology doesn’t always win—the technology that deploys with the lowest risk often does.” Is this the right part of the conversation to bring in that quote?

[00:30:44] Maribel Lopez: Sure, we can talk about it—it means a lot of things. Let me tell you the essence of that quote, because I think it’s important to understand. As a buyer, someone will come in and tell you, with complete confidence, that they have the best solution for whatever it is—”We have the best agentic framework for you.” Okay, that’s great—you might have the most technically advanced agentic framework, but does it work with everything else I have? Am I going to have to pull in an army of consultants for six to twelve months of integration work, MacGyvering something together with duct tape and glue just so I can have “the best”? And is “the best” even necessary, or is eighty-five percent good enough? Technology vendors hate it when I say this, but the reality is most of us are running our businesses on a tech stack that’s eighty-five percent good enough. Maybe we’re not running as effectively as we could be, but the wheels aren’t falling off the bus. So being great at what you do as a technology vendor is necessary but not sufficient—you also need the right partnering ecosystem, the right level of openness, and the right level of security and governance. There’s a lot you need to be truly deployable in any business environment. Especially as a startup, you tend to think it’s all about having amazing tech that works great.

[00:32:17] Scott Kinka: A hundred percent. I’ve said this multiple times on the podcast, and I’m curious how it lines up with what you’re saying: any tool that allegedly does everything, really doesn’t do anything particularly well.

[00:32:31] Maribel Lopez: That’s true, that’s true.

[00:32:33] Scott Kinka: Would you agree? You can’t be a master of all domains.

[00:32:37] Maribel Lopez: Right. And I do think you can be extremely effective at a lot of things, particularly if you’re a larger vendor.

[00:32:47] Maribel Lopez: But then there are always going to be tuck-ins—well, maybe “tuck-in” is more of an acquisition term, but there are always going to be things you should be partnering with or working with other organizations on. The reality is that everybody you or I talk to has a lot of technology already. And someone with a big AI platform—because having an AI platform is a thing in 2026—will roll in and tell you they have a great AI platform, and you’ll look at them and say, “Out of the fifteen things in your platform, I already have seven of them. How do we make this work?” Because you’re not ripping out the seven things you already put in—some of those you may have put in just last year, and your management definitely isn’t letting you rip them out. So the question is, how do you create a multi-vendor, arguably multi-cloud, hybrid AI cloud-and-on-prem strategy that’s manageable, has the least risk profile, and has a good security stack? That’s actually very complicated. It’s always been complicated—it just keeps getting more complicated as we go.

[00:34:03] Scott Kinka: Sure. Maribel, I could talk to you forever, but we’re going to be way over on my target for this episode. So let’s play a game—I’m going to hit you with a few questions, rapid-fire, so less than a sentence for each answer. Let’s see how we do. What’s one AI trend that you think is receiving far more attention than it deserves right now?

[00:34:27] Maribel Lopez: Agentic AI—because not everybody’s ready for everything to be autonomous.

[00:34:31] Scott Kinka: Love that. One question every board should be asking its CIO about AI that almost nobody asks?

[00:34:40] Maribel Lopez: What’s the security in this?

[00:34:42] Scott Kinka: Love that. What’s one leadership lesson you’ve learned from watching hundreds of organizations navigate digital transformation? You can make it specific to AI if you choose.

[00:34:55] Maribel Lopez: You need to be willing to experiment, but also willing to pivot fast if things don’t work.

[00:35:01] Scott Kinka: I love that. Twenty years from now, we’ll look back on this period—what do you think we’ll realize was completely misunderstood?

[00:35:10] Maribel Lopez: Completely misunderstood?

[00:35:11] Scott Kinka: Yeah, what did we get wrong?

[00:35:14] Maribel Lopez: We focused too much on it being AI, and not enough on what we were actually trying to do with it.

[00:35:19] Scott Kinka: I love it. Okay, good work on the rapid-fire round. Let’s have a little fun—just a couple more questions, and this time you can be longer than a sentence. What’s one bold prediction you have about enterprise AI over the next twelve months that you think will surprise people?

[00:35:35] Maribel Lopez: We’ll start to see more physical AI coming onto the scene.

[00:35:41] Scott Kinka: Give me an example of what you mean by physical AI, for folks who may not be following along.

[00:35:45] Maribel Lopez: Physical AI is—we’re going to talk more about how we can bring robotics and simulation into organizations. And simulation might just mean building your next building, as opposed to having robots in your corporation.

[00:35:58] Scott Kinka: I love that. What are you reading right now, on your end table?

[00:36:05] Maribel Lopez: I’m a little behind on my Louise Penny books—when I’m not doing intense AI work, I like to read cozy mysteries. So there you go, I’m reading The Grey Wolf right now.

[00:36:15] Scott Kinka: I love it. Tell us about your book.

[00:36:18] Maribel Lopez: My book was about how mobile data and IoT were going to allow us to contextualize experiences—and we never quite got there. Now, in the AI era, we think we’re actually going to get to personalization after all these years.

[00:36:32] Scott Kinka: So was there ever a follow-up to that book, do you think—or maybe a prologue to tie it into AI, or an epilogue?

[00:36:39] Maribel Lopez: What I say now is that everything’s moving so fast with AI that writing a book would be outdated within a month. So instead, we try to keep the blog, the LinkedIn newsletter, and the podcast alive to let everybody know what’s going on—just the way you are.

[00:36:54] Scott Kinka: How do people get ahold of that blog and newsletter if they enjoyed this conversation, which I’m sure they did, as much as I did?

[00:37:01] Maribel Lopez: Lopezresearch.com, /research, is where you can see all of what we’re writing about. There’s also a LinkedIn newsletter called AI Decoded—if you look up Maribel Lopez on LinkedIn, you can find me there and subscribe. I’ll leave you a link so you can put it in the show notes.

[00:37:20] Scott Kinka: Perfect, I love that. I have one fun question we’ve asked every guest for three years—it has nothing to do with AI, or maybe it does. It’s the end of the world—I don’t know if you’re a sci-fi fan, but you can pick whichever dystopian future you choose: zombie apocalypse, nuclear, whatever it is. Once this occurs, you’re alone, and you have one app still functioning on your phone, but you get to choose. Spotify? Your algorithm, tuned so well it just knows you?

[00:37:52] Maribel Lopez: You know what—if I’m all alone after the zombie apocalypse, I’ll have podcasts and music, millions and millions of songs I hadn’t heard, so at least I’ll be slightly entertained. And this is actually funny, after watching Umbrella Academy—I’ve been binge-watching Umbrella Academy for the past week or so, and there’s been a lot of apocalyptic discussion. So there you go.

[00:38:15] Scott Kinka: I love it—Spotify it is. That’s been a popular answer, by the way, over the last three years, so I get it, and I agree with you. One last question, and we’ll let you go: summarize your leadership philosophy in a sentence.

[00:38:29] Maribel Lopez: Allow people to do their best work.

[00:38:31] Scott Kinka: Words to live by—I love it. Maribel, thank you so much for your time.

[00:38:36] Maribel Lopez: Cheers.

[00:38:37] Scott Kinka: This has been amazing. That’s a wrap on this episode of the Bridgecast. If today’s conversation gave you something to think about, it certainly gave me something to think about. Make sure you hit subscribe, leave us a review, and share it with your team. We’ll add links to Maribel and her content in the show notes—super excited about it. Maribel, give us the name of your podcast.

[00:38:58] Maribel Lopez: AI with Maribel Lopez. Very easy.

[00:39:01] Scott Kinka: I love that—and certainly subscribe and follow that one as well. We’ll catch you soon on another episode of the Bridgecast. Thanks for your time.

Listen, if you made it this far, there must have been some good content, or you’re just a fan or a relative of mine. Either way, we appreciate you spending your most valuable asset with us—your time. We don’t take that lightly. We’d also like to say, one more time, that we’re appreciative of Bridgepoint Technologies and their belief in and sponsorship of this show. If you or someone you know is thinking about your company’s digital transformation, or simply the next IT project that you may not have the resources, budget, or time to get to, I’m sure Bridgepoint—one of the country’s fastest-growing technology advisory and procurement firms—can help. Check out bridgepointtechnologies.com—don’t forget the “e” on Bridgepoint—or simply reach out to me at skinka@bpt3.net. Also, please take a minute to give us a five-star review on your favorite platform—it helps give us the visibility to reach other people like you. Thanks for listening to this episode of the Bridgecast. I’m Scott Kinka, and until next time: there’s a lot of noise out there i