A Practical Conversation About AI in Business: From Hype to Real-World Impact

Hey, everybody. Welcome back to the show. I am your host as always, Ben Thomas. You know, one of the questions that we always get on Experts Talk is we talk a lot about marketing and b to b and best practices, but people ask us all the time about AI and how does this impact my business and how can I practically utilize it? What technologies are out there? And I said, you know what? I'm tired of people asking me this question, so I need to actually ask somebody who is a real expert, in this field to come in and answer those questions. And people keep asking me because I don't know the answers to them. So I said, okay. Well, let's do it. So that brings on, my guest today, Kelly Rascovich, executive editor for Deloitte's Deloitte's tech turns report. Kelly, thanks so much for putting up with me coming on the show today. Oh my gosh. Thank you for having me, and I would hunch you know way more than you think, sir. So know that I'll be asking you questions too. No. I'd I so I first of all, I'm the only person who can ask questions here. No. But it it is funny. Right? We we there is some level I feel like where where people do have more of an understanding of AI than I think that they really do, but people just get a little bit gun shy whenever you start to ask about business practices and ROI and how do I take something that I use in my personal life and deploy it at scale in a business practice. Right? They're just they're just some questions and and fears maybe that people have. And and look, you're the person you're gonna answer all the world's questions today. You cool with that? I'm cool. Well, Kelly, I I I'm curious. So really before we we dive in, obviously, I mentioned you're the executive editor of of Deloitte's Tech Trends report. That could need a lot of different things to different people. So can you break down specifically what what is included in that and what folks can expect whenever they inevitably read it at the bottom of this post? You bet. Okay. So tech trends is basically our technology point of view. So think about this as an annual report. We've been at this for seventeen years, and the intent is really to help distill that signal from noise. So really, what's new and next in tech? How can organizations start to think about this? Because at the end of the day, there are so many buzzwords kind of flying around and trying to understand what this means for me in my personal life, what this could mean for me in my organization is tough to grapple with. And so what we like to do is really ground ourselves in that notion of storytelling. You'll get great data as part of the report. And then you're also gonna get stories of this happening here and now. And a part of that is because the future is already here. It's just unevenly distributed. So this isn't just what's Deloitte's take on it. We are shining a light on globally, cross industry, how are folks already doing this so that you can learn from an oil and gas company just as much as you can from a consumer and retail one. So really meant to be that shining light of what is our technology point of view and how can you start thinking about this in really practical ways. Yeah. Well, I I appreciate you coming by sharing that context, and I know a lot of listeners and viewers of this show come from all sorts of different industries of b two b. Right? Oil and gas, you mentioned one of them, health care, hospitality, transportation, construction, engineering. And we're all sort of asking very similar questions because like I said, we're you know, many of us have GPT or APLLOWD or Gemini downloaded on our phones, and, you know, there are some legal requirements and hoops and stuff like that that maybe prevent us from using that in our our work world, But we're trying to figure out how how to deploy technologies like that at scale in a way that's compliant, meaningful, and really safe and secure as well. So I wanna ask this. Right? You know, I think it is maybe an oversimplification say there's a lot of noise in AI, right now. Right? You you see sort of conflicting reports maybe sometimes where I remember MIT came out with a a report that said, like, something like ninety percent of AI deployments and, you know, the enterprise have failed, which seemed pretty extreme, and then you hear other people saying all the opposite. Could you give us just, a quick baseline of, okay. Here's sort of where we are in the life cycle of AI deployments at scale in the enterprise, and then maybe we'll kinda use that as a launching off point to dig in a little bit deeper. Happy to. Well, first things first, in actually last year's report, a lot of the theming was around AIs everywhere. And so I think that's a really important kinda takeaway first and foremost because at the end of the day, what we're seeing is that really being table stakes and everything. So just like when we woke up this morning and turned on the lights, we didn't think about the electricity working or what is the h t t m l associated with our conversation today? Right? We this is just table stakes, and we're seeing that shift as it relates to AI too. And a lot of folks are still on the journey. Right? And at the end of the day, there's a lot to unpack here. And part of what we're seeing is that notion of moving from pilots and POCs to production and impact. And so there's a lot happening in this space. I know from a agent's perspective, when we talk about agents, thirty eight percent of organizations are piloting agents, but only eleven percent have them in production. And then let's even take a step back. Forty five percent have are figuring out what's their strategy associated with agents, but thirty seven percent have no strategy at all. So there's a lot here around how do we make sure we're leading with need, we're answering the right problems we're solving. And so at the end of the day, AI, I think we're on this journey. We're gonna continue to be on this journey, but we're really seeing that shift now from pilot to production in a meaningful way. But there's also a long ways to go too. And so I think there's beauty in that because we're all on this journey together. So learning from folks what's working and what's not is really helpful. You know, I'm curious. You talk about making that jump from pilot to production. What typically are some of those common inhibitors where people say, okay. We're ready to deploy this or we wanna try it. Right? And some of it is like, hey. We've got r and d departments who are trying this, and it's not ever easy see the light of day. But what typically are are some of those barriers to actual implementation from that that that that kind of trial environment to that production environment? So I think it's really easy for organizations let's think about this from like a cooking example, right, to focus on the ingredients, but really ignore the recipe. A stat we have in our report is that as we think about AI investment, ninety three percent of those AI investments are geared towards the actual tech, and only seven percent is the talent and people. And that ratio has to shift. And so I think that notion is when you're making cookies, you actually get all the ingredients in there. But if you're not following the recipe of when to take the cookies out of the oven, they're gonna burn. And so I think some notion of understanding how do we make sure that we're thinking about this holistically end to end. And at the end of the day, we can't have the tech ready, but the people not ready to. And so that ninety three and seven ratio, we've got to find a way to get that different over time because at the end of the day, trust Trump's tech, and you have to have people trusting the technology. You have to have the training and the tools. You can't just be talking about the model, the chips, the software. You know, it's interesting even that you you bring that up. Right? And and one of the things that I've seen a little bit more anecdote anecdotally with with some of, you know, my day jobs, some of my clients is there's sort of this fear that, like, we're gonna deploy something, and then tomorrow, it's gonna be obsolete. Right? And there is a legitimate like, that is a legitimate argument. Right? We are in in innovation cycles now that are minutes versus decades. Right? And how do you sort of approach that from from a a deployment standpoint? Right? How do you sort of start that process knowing that, look, two months from now, there may be something totally different or the or agenda may take this massive leap next week that we're not even ready for. How do you sort of balance sort of some of those conversations in this process as well? So to me, I like to think of it as a mindset shift, and there's a few things we're seeing leaders embody. So one, let's actually lead with what are the problems we need to solve versus the technology. Right? So let's solve let's lead with need. Let's understand how can we make an impact and really lead with those big business problems. I think the second part of that is really we have to tackle our biggest problems, not just those quick wins. Because I think there's an there's an inertia towards people doing small quick wins and expecting huge gains and huge benefits overnight as opposed to tackling those meaty, those hard problems, because those are really where you're gonna get that value unlock. And then I think there's a notion of how do you prioritize velocity over perfection? I'm an Indiana Hoosier alum, and they just won the national championship, which is great to see. And if I think about coach Signeti, right, like, he came into an organization and he prior he embodied that notion of, I'm gonna flip the script. I'm not gonna wait years to have the perfect team, the perfect playbook. I'm gonna try things, try it fast, try things in ways we haven't before. And also challenging that orthodoxy of just because we've done it a certain way doesn't mean we have to do it this way going forward. And so I think some of those mind shut mindset shifts are so important because that changes your culture, how you embody this stuff, and that also will really help you be on that journey of value of impact of ROI. And I think that velocity over perfection wasn't one is important because the pace of change is huge. So you have to get started somewhere. You can't wait to have all the answers. Well, I I love that you've been mentioned that. And and first of all, shout out to those, long suffering Hoosier fans like yourself. God bless you. You've you've earned it. I I love sort of attacking it from the end goal in mind. Right? And, a, not only does it help define sort of the ROI, but we fall into this trap. It feels like a lot of times as organizations of trying to stay new, fresh, relevant where it's like, oh, I just this is new and cool, and I wanna try it, and I wanna innovate. There's like, that's fine. Right? But it becomes more difficult to say, you know, oh, this new AI tool came out in our CRM. Let's use it. Well, okay. Well, sure. We can deploy it and use it, but what, like, what are we trying to accomplish? Are we trying to And to what end? Achieve, yes, shorter sales cycles? Are we trying to to achieve automation? Are we trying to deploy AgenTik for better lead gen? What does that look like? So I think I I think maybe some of the I don't know that frustration is the right word, but maybe some of the confusion, maybe some of the the, at times, apathy towards towards technologies and innovation might be because of their sort of aimless. Right? It's like, you know, I love I'll use an example. I love Suno. Right? But I sometimes I'll just go in there and create, and I get a little bit bored of it, and it's a great AI music tool. But I don't necessarily always work backwards from the end goal of saying, maybe I wanna create an album, or maybe I wanna have something done in six months. So for me, it's just fun and cool and not necessarily something that's actually practically driving a real tangible meaningful result. And I think there is a notion too of beyond just the, to your point, the tangible results, I think it's really important to design and build with people, not just for them. Because at the end of the day, like, co creation is so important, and people will endorse and use what they help build. I know Walmart revamped their scheduling app so people could change shifts and have that visibility into what their cycle is over the next few weeks. Right? And they built that with their employees. And so I think they saw that payoff in spades because at the end of the day, they designed that solution with their organization in mind, but with the people who are actually using the tool, and it was a success. Right? And so I think that notion of how do you make sure that you're designing with people, not just for them? Well, as we talk about how great people are, let's talk about not people. Let's talk about the assertive side of things. You touched on it a second ago, and it's one of the it's one of the the spaces where there's there's sort of just a lot of, like, okay. What? You know? Is this like a robot coming in to, like, do these things for me? AgenTex sort of is I don't know the topic du jour maybe is the right way to put it, but it's with sort of the next real meaningful kind of a a AI deployment deployments. But I think where it feels like, and correct me if I'm wrong on this, it really feels like that is really the step or the the the fighting line between where AI becomes this really cool nifty tool and actually starts acting more like almost an employee or a taskmaster or something like that. And that really like the agentic side feels like that's where you're gonna start seeing like scalable results and you know, because AI doesn't sleep and things like that. Do you do you see sort of agentic really being that next like, okay. This is this is where we're gonna be able to make these these meaningful business results. We see it, I think, as one, something that is taking off in a big way. And I think there is a notion of it's always going to be humans with employees. So we can absolutely talk tech, but at the end of the day, that humans with employees is or humans with technology is so important. From an agentic perspective, I think figuring out what does the Silicon workforce look like alongside your human workforce. There's some meaty questions that I think people have to solve to continue to be on this journey. Some examples. What does HR for agents look like? How do we hire an agent? How do we fire an agent? If if an employee trains an agent, and then that agent goes and deploys five other generations of agents, That agent that makes a mistake eventually, right, is that a is that an IT ticket that gets routed, or is that a performance management where they have to do mandatory training? Who owns that relationship? Who owns it? And who's responsible? Who what does the cost profile look like of an agent versus a personnel cost? Right? I know BMW is looking at agents as a you know, and seeing what is the cost profile from a human to having an agent. And so I think these are big meaty questions that folks are asking and rightfully so, because you have to have some of those answered before you get too far down the line on a lot of this stuff. I think there's another big takeaway here, which is we have to take a step back and redesign the process. Yeah. That might sound simple. That might sound table stakes. But if you slap an agent on a process that's not designed for agents in the first place, you're almost automating a broken process to begin with, and you're not gonna get the results you're looking for. And so I think taking that time to redesign versus automate is so critical. Well, look. I'll steal an analogy that we use here at Texas. I'm sure other people use it, but based upon a big is is what comes to mind when you know, I I I think that there's this idea that, like, AI is gonna come come in and solve every problem that I have is Yeah. Well, look. If your process is broken, you're just gonna do a broken process faster. Like, you know what mean? Yes. It's it's it doesn't come in and fix all your problems. But one of one of the challenges too, especially that that that I see pretty regularly in really hyper regulated industries that are, you know, had to be GDPR compliant, OSHA compliant, HIPAA compliant, things like that is that there's sort of this this. Okay. Yeah, this agentic side of things sounds really cool and it's good because I can manage a lot of the tasks and things like that. How do we take it really into that next generation of, okay, we're we're in security. We're in compliance. We're in, you know, these worlds that require real meaningful tangible training. How does like, where does where does that line sort of come in, and do you see, like, you know, maybe is it a more custom higher grade agent that comes in versus maybe, like, of an off the shelf? I mean, what what does that look like? I think it first and foremost starts with what are the problems we're trying to solve, how you want to use agents, and in what way. We spoke to HPE, and they said we have agent we are redefining what our operations and performance management looks like. Right? So they said, we are going to focus on a specific area and hone in on that. And we know that this is a problem we're solving. We can see this happen and really create those impact and value. And so I think honing in on what's the business case, how do you apply it, and have that focus towards where should you use an agent and where not. Because there's times, to your point, Ben, where you're gonna want you will need a human in the loop. So where where does it make sense to have an agent or not? If we think about cyber, I think at the end of the day, we one of our trends is all around the notion of fighting AI with AI because cyber, it's still a lot of the same playbook, but the speed and impact of which you have to now have this continuous oversight and update your strategy from a cyber perspective is really critical and really important. You can't wait to do you can't wait to have this be kind of an annual check or a quarterly or bimonthly check. This has to be continuous because the speed and impact in which AI is unfolding with this is really what's changing the game here. So I think kind of leading with need and then figuring out where do you want to deploy agents and how do you wanna deploy them. But giving yourself that permission to also take that step back and redefine is huge. You know, Kelly, this is usually the point in the interview where I start bringing up things like what's next, what people can expect. But but Yep. I I wanna totally flip that on its head. Right? Because it's it's easy for people like me who love to kinda poke around, and I'm sure people like you who just love sort of the knowledge and love understanding and trying new things that we we we sometimes forget that there are people at the very, very beginning of their journeys with this. Right? And people that that are not necessarily still writing on legal pads. Right? I mean, there's some of that. Right? But the heavy industrial, the agriculture, the manufacturing, you know, sometimes those industries tend to be a little bit slower to adopt things, and many of those people I know still from personal experience haven't deployed any sort of meaningful AI strategy at all. So instead of thinking future, where would you recommend that people actually start that conversation that haven't deployed anything? Agentic for them is like some crazy rough thing. Like, there you thought about it. Yep. I think it's starting with strategy and where are to having conversations with what are the pain points that you should tackle and solve. I had a close friend who called me this past fall and said, I reported to the CIO, and I need to do two GenAI use cases by the end of the year. And I said, well, hang on. Like, let's figure out what are your biggest problems in your organization, and let's start there. So let's build out what the strategy is in use case. And so I think it is taking that time to understand where you want to start and coming up with that strategy. Because otherwise, if you don't have a strategy and you don't have the tools to track ROI and to answer those meaty questions, I think you're not gonna get the results you might want or you might hope for. And so think at the end of the day, though, like, Ben, how cool that they're picking up their head and saying, let's start somewhere. And I think it's what's the art of the possible? How do you think about this holistically? And I love that. I know too from a trust perspective. We spoke about it a little earlier, but leaders trust AI three times more than employees. Right? And employees are probably the ones that know what the biggest problems are to begin with. So how do we flip that ratio and make sure that we are talking to the frontline employees that know what the real problems are and are living it day in and day out and solving those? Well, look, one of the things that I love to see, and and I I mentioned this in most of my episodes, like, this is by the way, this is not a paid sponsorship. This is not an ad placement. This is, like, purely editorial. One of the things that I've loved to see about, like, the consultant role that I'm specifically gonna pinpoint, Deloitte, is that I think maybe the the general consensus, we'll just say ten years ago, was that if I'm not this massive Fortune five hundred company, I would never even consider going to somebody like them for a large or for any deployment. You guys have done such a good job of saying, look, it's we we want to curate the conversations with the people talking about this the first time. We want to talk about some of those even in SMB use cases. It doesn't have to be this, you know, if you're not on, you know, the New York Stock Exchange, you're not relevant, right? I am so glad to see that Deloitte has done such a good job of embracing a lot of those communities and really just being a a great voice and teacher. How have you seen some of those those changes and that renewed focus be really beneficial for you guys? Well, one, thank you, Ben. And I think at the end of the day too, even if I think about the work that I do, a lot of what I do is connect with clients globally, but I'm also talking to academics, to thought leaders, to luminary, really opening the scope because at the end of the day, it's hard to see all the ingredients when you're inside the jar, right? So how do we have that holistic perspective? And I think that's the beauty in all of this is continuing to talk to the startups, partner. I know you said it, but I think success also looks like partnering with other folks that maybe you haven't in the past. I think it's gonna take an ecosystem of partners and players to tackle some of the things that folks wanna do. And so giving yourself that permission to ask what are the problems we should be solving, but then think through, is that something you wanna do directly within your own organization, or are there experts you can bring along the way to help curate the strategy or make connections that maybe we haven't in the past? Well, Kelly, look, I know that that a lot of people watching this right now sit across all sorts of different parts of that spectrum, and I know that there will be some folks who have some questions. What's the best way that people can reach out to you? Is it LinkedIn just to ask some of those, even maybe basic questions? Yeah. Feel free to reach out to me on LinkedIn. We you can read our report. I know I believe we'll link out to that too. But read the report, engage with it, and then any questions, feel free to reach out to me. I'm available on LinkedIn and love engaging with folks of what are they seeing in their day in and day out. Because to your point, we're all on this journey and at different points in time. So some of what I talk about might be tech trends from two or three years ago, but that's because it's relevant to that organization here and now. And then there's other folks who are, you know, way farther down the line. And so I think there's applicability here for everyone. And it's tech is influencing all of our lives in some way, shape, or form, and so really grateful to be on this journey with you all and learn from folks what they're seeing too in this space. Well, Kelly, look. A big thank you for me, not only for coming on the show, obviously, but but for your continued investment in these conversations. I think it's one of the the the biggest areas of confusion, and when you confuse, you lose. Right? And the more that we could teach and educate Yeah. People on the opportunities they have, ways that they can scale, even opportunities just to to make life a little bit more convenient for them. I think it's an important conversation from a business standpoint. So thank you so much for coming on. We we enjoyed your insights today. Awesome. Thank you so much for having me. I hope you have a great rest of your day. Well, thank you, Kelly, and thank you all for tuning in. Be sure to check us out next time on Experts Talk.