# From AI Experiments to Enterprise Value Driving Real Business ROI - with Dan Diasio of EY - **Show:** The AI in Business Podcast - **Published:** 24 Mar 2026 - **Source:** https://podcast.emerj.com/from-ai-experiments-to-enterprise-value-driving-real-business-roi-with-dan-diasio-of-ey - **Transcript:** https://podcasts.opennash.com/shows/ai-in-business/from-ai-experiments-to-enterprise-value-driving-real-business-roi-with-dan-diasi-66d8d9/ - **Provider:** OpenNash Podcasts (machine-generated) ## In brief Dan Diasio of EY argues that companies get the most from AI by redesigning how they work, not by automating today's processes. EY's AI Pulse survey suggests most firms see positive returns. Few use the gains to cut staff, and more reinvest in people. The biggest spenders shift from efficiency toward differentiation and product innovation. The main tension is between the easy path and the harder one. Buying tools leads to productivity and cost savings, but everyone uses the same models and gets similar results. Diasio says differentiation comes from people's mindset and skills and from rebuilding work end to end. Fixing one bottleneck just moves it elsewhere, so the redesign has to continue. ## Key insights 1. **Layoffs are loud, but reinvestment is more common** EY's survey found only about 17% of companies use AI gains to cut headcount, and more than twice as many reinvest in their people. Diasio says the layoff story gets louder coverage than the reinvestment story. 2. **Big AI spenders shift toward differentiation** Companies investing over $10 million reported less focus on employee productivity and operational efficiency over six months. They put more effort into competitive differentiation and product innovation. 3. **Automating today's processes builds yesterday's company** Diasio says you don't become AI first by automating what you do today. EY's approach moves from processes to 'jobs to be done', maps them to agents, and builds a new AI-native operating model. 4. **Shared AI models risk commoditized output** In EY's snack-company workshop, executives using the same models produced similar products and packaging. Diasio calls this a race to average and says human direction is what makes output distinctive. 5. **Fixing one bottleneck just moves it** Diasio's example is a drug company that speeds up discovery, only to find the bottleneck moves to clinical development. Leaders therefore have to redesign work end to end, one stage after another. ## Questions this episode answers ### Are companies using AI mainly to lay people off? According to EY's AI Pulse survey, as Dan Diasio describes it, only about 17% of companies use AI gains to cut headcount. More than twice as many are reinvesting the gains in their people. ### Is AI actually paying off financially for businesses? Diasio says 97% of surveyed executives report positive ROI (return on investment) from AI initiatives. 57% say the benefits are significant. The survey covers 500 US executives and is run every six months. ### What does it mean to be an 'AI-first' company? Diasio says it means rebuilding how work flows end to end, not bolting AI onto existing processes. EY breaks work into 'jobs to be done', assigns pieces to AI agents, and builds a new operating model around them. ### If every company uses the same AI models, how can anyone stand out? Diasio says the same models give similar results, which he calls a race to average. He argues differentiation comes from skilled people who direct the AI with original thinking: 'AI lifts the floor, but it's human ingenuity and people that lift the ceiling.' ### What should executives do to keep up with AI? Diasio says leaders should use current paid tools and not judge AI by older or free versions. He spends four to six hours every week staying current. He also suggests giving junior staff big problems, because they are less tied to the old way of working. ## Facts - **Matthew DeMello** [0:00:11]: "Today's guest is Dan Diasio, Global AI Consulting Leader and Americas Consulting CTO at EY." - **Dan Diasio** [0:07:00]: "we do every six months is we ask 500 executives from around the US" - **Dan Diasio** [0:07:27]: "97% say they're getting positive ROI from their AI initiatives, and 57% say that they're getting significant benefits from their initiatives." - **Dan Diasio** [0:03:15]: "our data suggests that only 17 of companies are using AI gains to be able to cut headcount" - **Dan Diasio** [0:03:15]: "For every one company cutting and and reducing their headcount, more than twice as many are reinvesting in their people." - **Dan Diasio** [0:09:15]: "major areas of focus, tend to be around employee productivity and operational cost savings" - **Dan Diasio** [0:10:13]: "there's been a significant drop in their focus on employee productivity and in terms of operational efficiencies" - **Dan Diasio** [0:08:18]: "those CEOs are no longer delegating it, they are starting to drive it" - **Dan Diasio** [0:16:57]: "In the early days, we realized that with executives, we could no longer teach them in PowerPoint." - **Dan Diasio** [0:17:21]: "we brought a computer in front of them, we gave them an objective. Chances are, that objective was gonna be impossible to do inside of 45 minutes" - **Dan Diasio** [0:17:21]: "One of the things that we had them do was to build out, um, a a plan to be able to create a snack company." - **Dan Diasio** [0:18:24]: "a lot of the packaging looks the same, a lot of the products look the same, they're using the same models, they're getting the same output." - **Dan Diasio** [0:21:27]: "the frontier of AI capability today is in the space of building software" - **Dan Diasio** [0:22:52]: "I have four to 6 hours a week I spend staying current with the technology" - **Dan Diasio** [0:23:26]: "We're talking in February 26, it is accelerating, it's been three months of acceleration, and it is continuing to go." - **Dan Diasio** [0:31:39]: "you are going to find this happens all the time that the bottleneck moves now, it moves from drug discovery into clinical development." - **Dan Diasio** [0:32:41]: "we have something we call our value blueprint. What we say is that we need to move away from processes to jobs to be done." - **Dan Diasio** [0:35:16]: "the computer scientist is now also an AI engineer and is also now a data engineer." - **Dan Diasio** [0:15:19]: "for many companies, they've invested in the technology stack over the course of the last two to three years." - **Matthew DeMello** [0:11:17]: "We had a guest on on the show not too long ago, David Carmona at at Microsoft" ## Transcript **[0:00:11] Matthew DeMello:** Welcome everyone to Emerges AI in Business podcast. Today's guest is Dan Diasio, Global AI Consulting Leader and Americas Consulting CTO at EY. Dan discusses the shift in enterprise strategy from using technology to automate existing tasks toward a total redesign of core business processes. He highlights how organizations investing over $10 million are moving away from simple efficiency gains to focus on how this technology enables entirely new ways of working and competitive differentiation. The conversation reveals that nearly all companies report positive financial returns, though few are using these gains to reduce staff levels. **[0:00:52] Matthew DeMello:** Instead, leaders are reinvesting those resources back into their teams to solve previously impossible problems and break through established operational constraints. Please note the opinions shared by Dan Diasio in this episode are his own and do not represent the official position of EY or its leadership. Before we begin, a quick note for our executive listeners. Emerging bias enterprise leaders who are driving meaningful AI initiatives to share what they are learning with the wider audience. If you're moving real projects forward and want to be part of that conversation, you can learn more at go dot merge dot com slash experts. **[0:01:30] Matthew DeMello:** That's G O dot E M E R J dot com slash E X P E R T. Now the conversation with Dan. Dan, welcome to the program. It's a great pleasure having you. It's a pleasure to be here, Matthew. Absolutely. Uh, you're here to talk about the new AI Pulse survey from EY. Before we dive into that survey itself, we're hearing a lot on the show about the growing tension we're hearing from senior leaders between treating AI as that kind of cost lever versus a strategic engine for differentiation and growth. **[0:02:05] Matthew DeMello:** And there's a big difference here. Uh, I know, uh, especially to new time listeners that might sound like a lot of buzzwords, but really i i is the i is we cover for so many different industries. This takes a lot of shapes, but a transformation, uh, as what happens to an organization adopting AI, is just that a transformation. Uh, you know, uh, the a lot of these organizations are are are uh caterpillars now, and they're gonna come out butterflies. **[0:02:29] Matthew DeMello:** But from EY's perspective with the survey, what is the most single counterintuitive finding, um, that you're getting from this effort? Especially where it shows senior executives tending to us uh, underestimate or misunderstand about critical aspects of AI adoption in enterprise today. **[0:02:47] Dan Diasio:** So Matthew, we are at a really interesting time right now. Because for the last three years, a lot of companies have been doing a lot of experimentation with AI. They're finding ways to be able to really power new processes, to empower people to do new work. But there's this conclusion that many of those companies are coming to, and that is you don't become an AI first organization by automating what you do today. That is essentially building and automating yesterday, not building for the future. **[0:03:15] Dan Diasio:** So, to your very specific question that you asked about what is the most counterintuitive piece of data that we got from the AI Pulse survey, I think it comes down to this. There's a ton of noise about layoffs in the market. But our data suggests that only 17 of companies are using AI gains to be able to cut headcount. For every one company cutting and and reducing their headcount, more than twice as many are reinvesting in their people. **[0:03:45] Dan Diasio:** And that layoff story is loud in the market. The reinvention and reinvestment story is a lot quieter. And we f like we've gotten that sense, but now we see it come through specifically in the data. **[0:03:57] Matthew DeMello:** absolutely. And I I think, um, I I think this is a point a lot of different industries are reaching. It's not just financial services, it's not even, I think, even delineated to the Fortune 100 or any kind of barometer you want to put on. Kind of the size of the organization, you can't differentiate this by, uh, SMBs necessarily to, uh, larger enterprises. But um, I think this this chorus that you're you're coming out with strong in your first answer. **[0:04:24] Matthew DeMello:** That the path to AI transformation or AI, an AI first organization is not going to be paved only in automation. I think is something that, uh, any listener, uh, across any industry can kind of hang their hat on. But um, I I really wanna nail down the formula here because it it does seem specific. And I don't know if this is necessarily everyone tuning into, uh, the show today as much as they might be, uh, educated by the rest of the conversation. **[0:04:52] Matthew DeMello:** But um, I I think there's only going to be, um, you know, so much of the. Uh. An area of the global economy where it makes sense to be AI first, to be, uh, you know, a completely fluid organization. I think we're at that point post the generative AI hype cycle right now, where, um, there was this, uh, very much, uh. An attitude and a chorus of, um, you don't want to wait to jump in the pool, but you want to jump in the deep end. **[0:05:17] Matthew DeMello:** And you want to tie it to business values, you want to, uh, make sure that this is goal oriented, you're doing something important. You're um s, you still m might have this in an experimental part, uh, of your business in a in a secure sandbox that's low risk, um, you know, high impact, um, but you're not working on anything that's unimportant. You're trying to do ten of an important job better, seems to be that kind of general formula, uh, for AI uh, success, especially in that, uh, that pilot phase. **[0:05:47] Matthew DeMello:** Moving on from the pilot phase, though, I I, I'm, I'm with you here, e especially as you're looking to bring it across the organization. Um, automation is a great place to start, but having that kind of uh, sort of pipe vision or tunnel vision focus on uh, automation alone, which is tempting in financial services, right? Cause there's so much paper flying around and paper's just a really great deterministic, uh, automation use case. Um, that, uh, if we really, really, really wanna take the next step, we have to think more deeply, and we have to think, uh, about how the the very way that our organization functions and how we're going to drive AI in all of those places. **[0:06:29] Matthew DeMello:** Um, just just for perspective there, uh, how do you have any examples, or do you have an idea of how? Maybe companies in various places of the global economy, uh, discover that they should be AI first? If that's maybe not how they started out in their journey, uh, with AI, maybe they just started on in experimental fashion, and then they go, you know, even despite what their competitors are doing, they're just looking at, oh, hey, to. **[0:06:54] Matthew DeMello:** To really make this business goal work with technology, we do have to be AI first. What does that look like? **[0:07:00] Dan Diasio:** Yeah, I think that's a that's a really great question, Matthew. And I if I if I even just back up from that, Please, um, you know, so so the AI pulse survey, the statistics that I just ran off there, I mean, so what we do every six months is we ask 500 executives from around the US. Uh, what? And how they are doing with their AI investments? Are they investing more in a particular space? Are they investing in, uh, in in building out their capabilities? **[0:07:27] Dan Diasio:** What are the blockers that are holding them back? And we always add in a couple of nuanced questions that are specific for that survey. So this this time, we asked them the question of what they're doing with their savings, that they're getting. Because most of those companies report that they are getting positive ROI, 97% say they're getting positive ROI from their AI initiatives, and 57% say that they're getting significant benefits from their initiatives. **[0:07:52] Dan Diasio:** So the question is, what are you doing with those benefits? And that's where I started to say the most counterintuitive thing is it's all about job reduction or layoffs. But only 70% said they're doing that. Actually, 38% said they're reinvesting it in their people. But I think, you know, just to get it back to your question, the data tells us that everybody is in on this AI journey. Like there are. There are very few companies around the world that have decided that AI isn't applicable to them at this point in time. **[0:08:18] Dan Diasio:** Yeah, and I think two years or three years of experimentation has led to a pivot that we're seeing in the market now, where it's no longer. The CEO is delegating that particular, uh, an AI initiative down to the CIO or the CTO, or the chief strategy officer. But the CEOs of the companies that are really starting to shift from bolting AI onto the way they work to building it into their core processes, those CEOs are no longer delegating it, they are starting to drive it. **[0:08:48] Dan Diasio:** So that is like a a pivot, like it's gone up upstairs even further inside these organizations. That we need to start to shape the future of what our company looks like, whether it's changing the business model or whether it's starting to really reinvent the operating model. The data that we have also suggests some pretty interesting pivots. So I wanna share another another interesting data point that we saw. hit me, So we see so we ask companies, how much are they investing in their AI initiatives? **[0:09:15] Dan Diasio:** Are they investing more than 500, zero? are they investing more than a million? are they investing more than 10 million? And it helps us be able to segment the data in a variety of different ways. So those that are over that, those that are investing $1 million or over. The areas that they are getting the bow the most amount of um benefits, uh, from their AI initiatives, like their their major areas of focus, tend to be around employee productivity and operational cost savings. **[0:09:41] Dan Diasio:** So if you invest some money, you're and you and you often think about this as a technology, then you expect that that technology will then create value. The way it creates value is it does what some people do today. It does the as you've described earlier, Matthew, the ten or the twenty of your job. And then from that, it starts to, you know, translate its way into some sort of economic benefit. The trick is that most of our clients, many of our c, many companies have really struggled in taking that ten or twenty savings of a job and turning that into value. **[0:10:13] Dan Diasio:** That shows up in the P. L. But if you invest more than $10 million, that's where the data starts to get different. It's actually over the last six months, there's been a significant drop in their focus on employee productivity and in terms of operational efficiencies. And instead, those companies that are investing more are putting more effort towards competitive differentiation and product innovation. So we, you know, like, there's this, there's this narrative, and I've even heard it come with some of your podca uh podcast guests in the past. **[0:10:44] Dan Diasio:** That like, we need to shift away from what we do today to shifting to what we could do if a lot more things were possible than what we. What we were previously constrained with, and that is starting to take hold in those companies that are shifting towards leadership. They're they're not thinking, you know, getting caught in something I call the visibility trap, where they look at their processes today and they figure out, how can I make AI within my processes work? **[0:11:08] Dan Diasio:** And instead, they say, I'm gonna throw out the old processes and I'm gonna try to create an entirely new way of working because I can now do a lot more than what was ever possible in the past. **[0:11:17] Matthew DeMello:** uh, that makes a lot of sense. I I also wonder, like, um, especially for for your survey. We we had a guest on on the show not too long ago, David Carmona at at Microsoft, he he, you know, he was at the ground floor in like, 20 years ago, uh, at Microsoft, really putting the foundations of AI into this historic organization. And he was telling us he's he's in the R D space now, and he's telling us, you know, uh, so much of AI adoption, especially, and this is true of financial services most of all, right? **[0:11:48] Matthew DeMello:** So much of AI d adoption, especially for the goals from the C-suite. And leadership has always been cost efficiency, cost efficiency, cost efficiency, less less. And that makes perfect sense. We know why, and that's a great use case in in many ways, or or is a great, uh uh, an enormous category for so many wonderful use cases. Um, especially, uh, really strong beachhead use cases that are great for starting an AI adoption process. **[0:12:14] Matthew DeMello:** Put putting, uh, automation in a very simple, deterministic place, but also a place, uh, where you can put technology and start to drink in really important data. That that, uh, tells you more about how your organization functions overall vis a vis your vis uh, business goals. But he also mentioned that well, this this sort of tunnel vision focus on cost efficiencies un tends to undermine the argument that, uh, you know, AI, and this is specific to R- uh, pharma, R D, or in R D largely, we can just it. **[0:12:44] Matthew DeMello:** It might cost more, it might not be as cheaper, but we can run better experiments with better data. And um, and we can do things, uh, like, uh, have, um, better simulations. If we move the conversation away from this constant focus necessarily on cost savings, cost savings, cost savings, uh, cost efficiency, cost efficiency, cost efficiency. And I know a lot of, uh, your last answer is focused on reinvestment. But I'm also wondering, is there a stomach you're finding in the survey among leaders across industries? **[0:13:19] Dan Diasio:** That's maybe somewhere specific where they're not just thinking about AI to do, uh, to do less. They're not deploying AI to do, uh, n something necessarily faster, but better, something more that requires more investment, but that's better tied to their business goals, is gonna bring them more. L revenue in the long run because they're providing a better overall product. Are we seeing anywhere, uh uh, for for that, or are we seeing kind of a large scale sense of this tunnel vision? **[0:13:50] Dan Diasio:** That, you know, uh, you, you best place to start for obvious reasons is, you know, bringing down costs, bringing down costs, bringing down costs. Yeah, I mean, I think when we start to segment the data between th- those that are investing more and those that you know, those that are investing more than 10 million and those that are investing more than a million. You see that the points of focus exa- uh, you know, get exactly to what you're saying, Matthew. **[0:14:15] Dan Diasio:** Like, if you're investing in product innovation, you are investing in growth. If you're investing in competitive differentiation, you are likely investing in your business model. So that so we are starting to see that sense. But back to your point about your uh, about about some of the observations, can I, can I turn the tables and ask you a question? please, I think **[0:14:35] Matthew DeMello:** yeah, actually, that this is refreshing. Hit me, go ahead, **[0:14:39] Dan Diasio:** all right, so so Matthew, I want you to give me a one word answer. **[0:14:42] Matthew DeMello:** all right. **[0:14:42] Dan Diasio:** I'm gonna ask you, what is the first word you think of when you think of AI? **[0:14:47] Matthew DeMello:** Well, that's loaded being in this job, but I'm gonna keep it to real. Rochek test. And I'm going to say, Uh, just because I was rambling on a second about it ago and I'm on one cup of coffee. Efficiency, Efficiency? Okay, I'm part of the problem. **[0:15:00] Dan Diasio:** all right, good. **[0:15:01] Matthew DeMello:** See, there's my my tunnel **[0:15:02] Dan Diasio:** Well, well, I mean, you, you actually went in a different space than most people go. I ask that question a lot, and most people give the name of a technology, the technology that they use as their **[0:15:13] Matthew DeMello:** It make makes sense cause I and cover a lot of different d different products, so so the one might not stick you, you have out to me. **[0:15:19] Dan Diasio:** a you have, you have a breath. You get to meet with a lot of different people. This is clearly a topic that is that is on your mind. Um, but you know, for many, for many people, they think of technology, and if you think about technology, then and and you, that's. There's probably been a lot of investment, like for many companies, they've invested in the technology stack over the course of the last two to three years. **[0:15:41] Dan Diasio:** They've invested in bringing in the right tools and then making sure that people had availability of those tools. Whether it be a platform, whether it be an agent orchestration capability, whether it be a garden of different models that our clients can use to be able to kind of pick the best one at any particular point in time. And that investment in technology means, you know, eventually you're gonna be asked the question of what your return is for that. **[0:16:04] Dan Diasio:** And when you invest in technology, you go right down the path that you were going, Matthew. Technology creates efficiency, so, you know, it puts you down this path of cost savings. And you're asked to justify to your CFO or to your board. What are the inc improvements in productivity and cost savings that you see? What we see our company shifting away like it actually takes three, three different kind of capabilities to drive a different looking organization. **[0:16:30] Dan Diasio:** You have to have the right mindset with your leaders and your people, you have to have the right skill set with your people to be able to work with this technology, and you have to have the right tool set. And most of, and many companies have been investing in the tool set first and then getting to the mindset and the skill set. And we see that starting to flip now, after all that focus on the on the technology that puts you down a path of driving uh, operational efficiency and cost savings in the P and L. **[0:16:57] Dan Diasio:** Now, if you shift on the mindset and the skill set, you can start to open up the conversation towards competitive differentiation. And and that's really important in this space because we're all using the same models. I mean, we have this example that I've used, uh, that, you know, this story that that we've told in the past. Um, where we, we. In the early days, we realized that with executives, we could no longer teach them in PowerPoint. **[0:17:21] Dan Diasio:** We had to start to create a real experience where we brought a computer in front of them, we gave them an objective. Chances are, that objective was gonna be impossible to do inside of 45 minutes, and we got them working on it. And usually it involved, like a a very scripted narrative. One of the things that we had them do was to build out, um, a a plan to be able to create a snack company. And um, uh, they had to look at friends, you know, cause everybody knows about snacks, you know, some people snack more **[0:17:49] Matthew DeMello:** Not not everybody might have a a a an iPhone necessarily, but everybody has had a Hershey's kiss in their life. Yes, **[0:17:57] Dan Diasio:** everybody knows about snack. So so we gave them a bunch of data of what categories were growing and not, and just said, all right, you need to create a recipe, a product, a packaging, you know, a a a sales and marketing plan, maybe a jingle, maybe a video. You have access to these tools, you have 45 minutes. And it was fascinating when you get the when, you know, I think there's. There's a lot that many executives need to do to be able to get their hands on the keyboard and actually feel what is happening to us these days. **[0:18:24] Dan Diasio:** So, um, so this was an exercise to be able to make that, and you, you know, it was a. It was a magical experience. You hear people laughing, joking around, um, and they would pitch their product out to to their peers. They'd work in small groups. But what we started to see after we'd done this many, many times before is that a lot of the packaging looks the same, a lot of the products look the same, they're using the same models, they're getting the same output. **[0:18:48] Dan Diasio:** AI made it incredibly easy to be able to get to a good output, but it commoditized that output. Now, what it takes, like, you know, now, what, like that is, um, that is like a regression to the average, uh, like something that we call the sammy. It's a race to average. So so AI proliferating around the organization actually needs to be human powered in a way. It's the human ingenuity that directs the intent for the system to be able to create something that is differentiated. **[0:19:20] Dan Diasio:** That is investing in the in the mindset and the skill set of our people. Because it's not just now, you know, instead of asking and deferring to this authoritative voice of AI, now we need to direct it with the appropriate, um, the appropriate thinking. Before we jump into the experience, we we like, say that the loop needs to be from a human to the AI system, to a human. So I think as companies are starting to invest more in this capability, they realize that they can start to unlock things that really drive differentiation and allow them to be able to create opportunities for greater growth or greater connectivity with their customers. **[0:19:58] Matthew DeMello:** absolutely. And and and just because we're we're this far out from the I'd call it the peak of the hype cycle with generative AI, we're just about three years out. Really, really hit a crest in in my mind around late 2023, early 2024. So maybe maybe we're a little bit more like two years out. We're definitely still c climbing up the roller coaster, uh, of the hype cycle and the understanding cycle, uh, for new agentic technologies. **[0:20:25] Matthew DeMello:** Um, but what advice do you have for leaders, especially the ones you're seeing in the survey? Um, about just being able to maybe m uh, address the misconceptions, um, the popular misconceptions that the these sort of zombie facts or myths that seem to be about these technologies. Um, e even even generative. Now that we're when we would expect the the truth would be out by now, you know, hey, we know we're not, uh, we, we're not gonna get a a two week uh uh version fix, um, fro uh from Silicon Valley. **[0:21:01] Matthew DeMello:** That just snap of our fingers fixes all hallucinations. I remember that was a thing, um, when I when I used to go on LinkedIn, uh, back in 2023, there'd be some guy from Silicon Valley saying, Oh, we'll have, we'll have hallucinations figured out in a couple months. No, no, it like that just fundamentally misunderstands w uh, what you're dealing with. Uh, are you s uh? W What zombie myths would you try to address, maybe with this audience off the bat, from what you're seeing in the survey? **[0:21:27] Dan Diasio:** So Matthew, I I love the question, and I frame it a little bit. Um, I I give you. Imagine you had the opportunity to invest in two different companies, one of those both of those companies have the same AI capability, and the frontier of AI capability today is in the space of building software. So let's say there are two companies. You have one that uses AI to build its software inside its organization, so it doesn't need as many software developers. **[0:21:54] Dan Diasio:** So it basically has now done this at a lower cost front cost footprint and is more efficient. The other one has taught their software engineers how to create ten times more software. And instead of thinking about a limit to the amount of software that they thought they could do and like getting stuck to their product backlog, instead, you've unleased that team to be able to go build what was previously impossible. If you as an investor had to take a bet between those two companies, it it's it's, you know, it's not really even, um, a hard it's not a hard decision to make, it's not a trick question. **[0:22:25] Dan Diasio:** You'd bet with the company that thinks that this technology creates abundance. So, you know, I think the myth is, how do you create differentiation? And the differentiation is really deploying your workforce in a way that goes and tackles what was previously impossible. So many organizations have worked with a set of constraints that were constraints around human capital. And now you can start to press against those constraints and start to really challenge at a a a point of something entirely different. **[0:22:52] Dan Diasio:** So that that's one thing that I think is just like a a helpful frame to to be able to, um, to apply as you're as companies are are really trying to navigate these these very bumpy waters. The second point that I think is really important is companies, and in particular executives, need to find time to invest in staying up to speed. Like I spend four hours to 6 hours every week, I have a, you know, I have a crazy schedule, but I have four to 6 hours a week I spend staying current with the technology. **[0:23:26] Dan Diasio:** The technology has not stopped, you know, it has not plateaued, it is actually accelerating now at this point in time. We're talking in February 26, it is accelerating, it's been three months of acceleration, and it is continuing to go. So, if your experience is in working with an AI system from last year, where you had a hallucination, you said, you know, like this isn't all cracked up, w uh, like this isn't all it's cracked up to be. **[0:23:50] Dan Diasio:** Actually, you have to work with the technology of today and not the free versions. Like, you know, I'm not advocating that everybody go out and spend money with these different technology providers, but find a way to get closer to the frontier. Because then you will stay informed of what's coming. A around some of these corners. And you'll be better in B- in a better position to be able to advise the direction of where you can apply that capability inside the organization. **[0:24:14] Matthew DeMello:** so so really, uh, just to maybe put a little bit of a nutshell on on what you're trying to say here is I, uh, I, especially to focus on the zombie myth. There's, uh, there's something of a myth, maybe. I it's it's s um or or kind of a a a an unspoken misunderstanding that we're kinda seeing out there that the different that the technology is gonna give you the differentiation. Uh, no, Dorothy, uh. The differentiation lives inside you, lives inside of your organization. **[0:24:41] Matthew DeMello:** But it might not be brought out with the technology and the uh, without the technology. And that is an important distinction to make. It is, uh, the but the it that the technology is not the reason that you will differentiate yourself from your competitor. But more that that differentiation, um, can be accented, can be strategically planned for, can be executed in a way. W uh, that is distinct, um, to, uh it, the unique position, uh, that your product or your service is is bringing to the market. **[0:25:11] Matthew DeMello:** Am I getting that sense of of maybe the more, uh, contrarian take you're having on, uh, the way we talk about differentiation and AI adoption? Am I am I getting that down? Uh, at at at all, just in that summary. **[0:25:24] Dan Diasio:** Yeah, Matthew, I would say that AI lifts the floor, but it's human ingenuity and people that lift the ceiling, right? And um, and we need to make sure that we are figuring out how we are applying, staying on the frontier to apply that capability to unlock the real potential of human ingenuity. And the example in the example that you described with, uh, with the, uh, the pharmaceutical, uh, example, R and D and pharma. I mean, they have been constrained on the number of different possible drug permutations they can come up with. **[0:25:55] Dan Diasio:** If you can, if you can apply AI to be able to go explore a lot more further, that is gonna be directed by some of your deepest scientists. But then you can start to unlock possible drug combinations that you never would have been able to prioritize getting around to explore. So it's so it's that, you know, AI is just making everybody better, and but it's not, it's it's people and teaching our people how to really challenge what's been possible in the past, um, or challenge what's been impossible in the past. **[0:26:25] Dan Diasio:** That is what drives the differentiation in the market. **[0:26:28] Matthew DeMello:** uh. I. I'm starting to remember which conversation that was. And full disclosure, I won't promote it anymore, cause now I'm realizing it's one of, uh, e EY's competitors, so we'll leave it at that. But the chorus we hear from everyone in R and D, R and D across life sciences, period, no matter what the logo is, uh, is that these technologies are so awesome. Um, uh, not even in the 80s sense of slang sense of the word, but really in its in its, uh, pure, uh, uh, astounding nature, uh, it. **[0:26:59] Matthew DeMello:** These technologies are s are at a scale that so collapses our, our typical understanding of what's possible. That we can come up with more, uh, more cures for rare diseases than there are grains of sand on the planet. Is tends to be the line, uh, that, uh, that I came up with, uh, with with with another guest, uh, from a podcast. To really help folks at home wrap their heads around how far we can take, uh, these technologies. **[0:27:27] Matthew DeMello:** But then on the business side, just as what you were saying before about, uh, life sciences leaders, uh, historically being constrained, uh, by the number of of uh treatments that they can develop for for certain, uh, conditions. Now that that metric is completely, it might as well be an. It's not infinity, it is finite, but for the human mind and for the amount and that's possible that we can achieve, it might as well be infinite. We can come up with more, uh, cures for rare diseases than there are grains of sand on the planet. **[0:27:55] Matthew DeMello:** There's still, uh, there's still a problem with business execution, because you're not gonna bring even the most r ridiculously small fraction of that to the market anytime soon, given the status quo, even for how fast the technology's moving right now. So then the question becomes, all right, for this awesome power you have on the input side, how can we best manage outputs to, uh, to meet, um, the the the focus required to even make good of of the, uh, you know, astounding possibilities. **[0:28:28] Matthew DeMello:** Um, uh, of where we could take this technology into the future, those are, uh, just because you can come up with all those solutions doesn't make, uh, the uh, these solutions come out any faster, or or bring them, uh, or or or cure lives in and of themselves, right? Just cause you found all of, uh, all of these solutions. So I, I, I think that definitely illustrates a lot of what you're saying around AI. Really, just isn't. **[0:28:53] Matthew DeMello:** It is isn't just, oh, hey, I have the superpower now. Oh, hey, I can, you know, fly a across buildings. I have this awesome, tremendous capability. It, it's about what practical good are we using it for. Great that you can fly. Can you save the cat in the tree? Uh, to keep this in the Superman metaphor, I'm I'm sort of coming up with right now. Um, and and and just to to kind of wrap, I know we've talked a little, we've kind of jumped around our little outline that we agreed on, uh, before we turned, uh, the recorders on. **[0:29:21] Matthew DeMello:** But um, last place here, um, uh, I I think the the one place we haven't really touched on is, uh, productivity gains vis a vis workforce strategy. Because, uh, I I think now we're at the point of the conversation where we're firmly outside of the hype cycle. Talking points, um, that I find are in kind of the zero sum fallacy, um, that that's out there. This is, this is my big zombie myth that I that I don't, uh, think does anybody a lot of good. **[0:29:47] Matthew DeMello:** But there seems to be this zero sum fallacy still out there, uh, on both sides, both labor and management. Where labor is saying, Oh God, um, they're gonna hire x amount of robots and get rid of y amount of human beings, and then we're all gonna be out of a job eventually, trends being the same. And then the M the management side, in this very cartoonish view, says, Oh, great, I'm gonna be able to hire x robots and not have to hire, or I'll just be able to fire y amount of human beings. **[0:30:14] Matthew DeMello:** And any time I've had anybody on the show talking about their process, uh, not even for the sake of politeness or political correctness. Does it ever, ever sound anything remotely like that zero sum trade off? It is so much more nuanced for how you're developing these systems, for what do you want them to do? And then you get to these points, as we've had in this conversation, where you find human input is inseparable from the awesome, uh, capabilities, the ability to see information in real time, the ability to to, uh, to track deep insights that no human being would ever be able to see. **[0:30:51] Matthew DeMello:** Uh, no matter how much time, infinite amount of time you gave them with spreadsheets or IQ pills, so to speak, um, you know, uh. But you find these capabilities become inseparable. That awesome real time AI, uh, technology that you've deployed in the workplace is useless without human oversight, without human, uh, inputs. To make sure that they're that the that, uh, the impact on the product is gonna be safe for other human beings to use and vice versa. **[0:31:18] Matthew DeMello:** Um, just very interested on what this means for workforce strategy. How are leaders thinking about this, especially once they get past that sort of zero th- uh, zero sum zombie myth I I mentioned a moment ago? **[0:31:32] Dan Diasio:** So, Matthew, I'm gonna definitely get to Please, your question, but there's something that you alluded to that I just wanna pull the thread yeah, go for a little bit **[0:31:39] Matthew DeMello:** it, go for it. **[0:31:39] Dan Diasio:** more. So, so we were using this fictitious, you know, uh, example of a company that is a pharmaceutical company that can reinvent the way they do drug discovery. And then you are going to find this happens all the time that the bottleneck moves now, it moves from drug discovery into clinical development. So, you know, a process when you start to think about things from an end to end perspective, an end to end process that is, uh, AI native, AI native, manual, and then AI native is a manual process. **[0:32:11] Dan Diasio:** You have, like, until you think about it from an end to end perspective, these things don't really transform or start to unlock something that is AI native. The thing holding many companies back, like in this case in the pharmaceutical, we'll get into clinical development. You actually need to start to think about what does the entirely new way of doing clinical development look like? And it's, it's, it's, it shifts the idea from a project to a transformation, shifts from a technology transformation to a wholesale business transformation. **[0:32:41] Dan Diasio:** And in order to do that, you need to reinvent the entire flow of the work. It's not about finding the seventy five processes that we used to follow to be able to do clinical development and now figuring out where the biggest AI opportunities are. But it's, you know, what we say at EY, we have something we call our value blueprint. What we say is that we need to move away from processes to jobs to be done. **[0:33:02] Dan Diasio:** We map those jobs to be done out to different things that agents can do. And then we rebuild in a whole new layer of what the operating model for the company looks like that is much more AI native. And these are the steps that are taking. Once you once you solve it in one particular part of the business, then you move the bottleneck to the next. And you just kinda, kinda gotta work your way through and start to think about how you build this on a more end to end basis. **[0:33:26] Dan Diasio:** So I just wanted to draw that point because that is the single thing that is holding our clients back from this idea of shifting from like a a bolt on company to a built in company, or a an AI assisted company to an AI first company is figuring out that AI native operating model. But but to your workforce, um, you know, there's there's something interesting that's happening in the workforce right now, where a and this is. **[0:33:51] Dan Diasio:** This is so unfortunate. Is that many college graduates are having a difficult time getting that's a a job. big one. Yep, Because I mean, it's, it's, it's real, like I'm on campus with many, uh, with many universities, and I hear this. You know, some of the some of the computer engineering students, like the the the people that go through those programs, are having the most difficult time in getting jobs. And, um, you know, often what we find is that if you really wanna think about something entirely different, you can't be too anchored into your past. **[0:34:21] Dan Diasio:** Like sometimes, your most valuable asset is the person who has not been doing this thing, doing doing work a certain way for the last 20 years. And some of those graduates are actually we find in our organization. They bring this disconnection from the status quo and start to because they've been growing up natively with these technologies over the last couple of years. They start to question, why would you do any of the work the way that you do it today? **[0:34:46] Dan Diasio:** So I think, you know it, you know, back to the point that you were mentioning about, uh, misconceptions. You know, there is a notion right now that people that have deep domain expertise are those that are most valuable, and I think they are incredibly valuable. Cause they can steer the direction of what's good and what is what is good, AI taste and what good output looks like. But when it comes to start to reinvent the business, sometimes some of those most junior folks are the ones that can actually help you make the leap forward and break you from the binds of your status. **[0:35:16] Dan Diasio:** So we incorporate this into our workforce strategy. We we really value some of those youngest people and put them onto big, hairy problems where we ask them to start to challenge the way we've done work in the past. And what we're finding, the, uh, the last point I'll make on workforce strategy is we're finding that a lot of these things are converging. So the computer scientist is now also an AI engineer and is also now a data engineer. **[0:35:40] Dan Diasio:** And they need to be much more of a product owner and have that product owner mindset. And that person coming together is able to help direct these systems to be able to do things that used to require teams of people working together over months to be able to achieve. **[0:35:54] Matthew DeMello:** A and you see it over and over and over again. The part of the business that was the most ignored ends up being going from a cost center to a revenue center. The employees at the business that were kinda treated like a joke cause they were just glorified mailmen and interns actually are going to be. Once you put them in the driver's seat of an AI adoption because they're the real subject matter experts in the workflow. **[0:36:18] Matthew DeMello:** Um, even if you never saw it that way before, um, you know, they're gonna, that's, they're gonna be the market differentiator for you. You just gotta deploy the awesome technology in the right way and and tune into shows like this with with smart guys like you. Uh, to to give us the advice that that, um, you know, h helps us get there as as an organization. It's those places where, up until this point, we d we didn't think this was important every step of that way. **[0:36:44] Matthew DeMello:** You, uh, uh, of of these hype cycles, of these technology adoption cycles, it's it's always those points and always those factors, always those workers, always those workflows that end up being the differentiator for the next step of the process. Uh, those are the keys. You're what looks like trash is, uh, so to speak, and and when I'm talking about human beings, please take that with a grain of salt. What might look like refuse, what might look like derelict is actually your treasure, uh, especially when you get to that next phase of process. **[0:37:14] Matthew DeMello:** And I can't tell you how many times, uh, I've noticed that across industries and with guests on the show. I know we've had a lot of your time today. Really appreciate you sticking around for a couple more minutes and really diving deep onto this, uh uh, onto this survey. I know we only covered so much, so if you wanna come back on and talk about even more, um, very happy to have a part two for this. **[0:37:32] Matthew DeMello:** But Dan, really appreciate you being on the show this week. Really fascinating stuff here. My pleasure, Matthew. Had a lot of fun and, uh, thank you for the invitation. Wrapping up today's episode, I think there are three key takeaways from our conversation with Dan Diasio, Global AI Consulting Leader and Americas Consulting CTO at EY. First, we saw that true organizational transformation requires moving beyond automating existing tasks to redesigning core business processes from an end to end perspective. **[0:38:01] Matthew DeMello:** Second, while lower level investments often focus on immediate efficiency, larger strategic commitments allow leaders to prioritize competitive differentiation and product innovation. And finally, the most successful strategies treat personnel as the primary driver of ingenuity, reinvesting efficiency gains back into the workforce rather than using technology to reduce headcount. Position your brand alongside the Fortune 500 leaders defining the enterprise AI roadmap for the opportunity to showcase your solution to the executives currently funding and scaling global initiatives. Partner with Emerj Secure your partnership at go dot merge dot com slash partner, that's g o dot e m e r j dot com slash p a r t n e r. **[0:38:48] Matthew DeMello:** For further executive level analysis and to join our network of leaders delivering workflow impact with AI, visit Emerj dot com. On behalf of the team at Emerj, we'll see you on the next episode.