A (2:33)

Yeah, it's really hard to, with any set of confidence, really predict where we're going to be in 2030. The way that we think about it is that when you have this amount of uncertainty, or, you know, a term that's often used is vuca, then the best thing to do is to be able to do scenario plans. And in this instance, what we did was we created four different scenarios or four different worlds that will be in 2030. And then you can start to plot to see how you can win in each of those scenarios. So on one end of the continuum, we have a scenario where, for a variety of reasons, AI of the future doesn't look that fundamentally different than the way that it does today. You know, you could have had supply chain, supply side disruptions. You know, there could be a bubble that pops, or there could be increased regulation due to some very public failures. But the AI doesn't look that much different fundamentally. And on the other end of the continuum you have, there's this very powerful AI that is really consolidated, where power is centralized into a single company or a single country. And if you're a company that's really trying to win in this space, you don't really have your own AI. You're renting it and your landlord is in complete control of how you pr, how like, you know, what you pay for that. So it's a, it's kind of an interesting, you know, thought experiment because many times if we just spend time listening to the, to the news, you know, we all kind of think that one future is happening. You know, it's just, you know, and like in different communities there are different sex. Like there's the, there's the group that thinks a little bit more about, you know, the AGI, asi. There's some that say it's just going to be agentic AI. There's some that are saying, like, this is all just a little bit of hype. So plotting the course of how the business really can win in each of those scenarios is the, is the game that we made for all the participants. And in that way, it was a lot of fun. They worked in groups of 10, they competed, they had options, they presented back and forth to each other, and then we awarded them some prizes. So. Fun way to spend two hours. Nice.

B (4:45)

So what did the, when you looked at the winning groups or the ones that were most prepared for multiple futures, what did that look like? What were the defining characteristics?

A (4:56)

So there were a couple of, really, I'm glad you asked that Question, Jeff. There were a couple of really interesting observations that we had. The first thing I would say is that I heard from many people afterwards that we had the right answer and then we asked our AI assistant and it gave us the wrong answer and then we changed our answer at the last minute. So that was kind of fascinating because we'll probably spend some time talking about some. Sometimes over relying on these systems can put you down a path of predictability and sameness that prevents you from really being able to unlock its full potential. Some of the things that we saw from the teams that won, they had a couple of creatives around the table and they were able to articulate and vision a new story of business model innovation or a new product that they become over time. And we're able to explain that in a really succinct and compelling way. So, you know, just one of the reflections I had is like the real piece of value in this space of differentiation when everybody is normalizing around the same capability, is the creativity that companies are able to or that in this case, that teams are able to unleash. So that would probably be one of the big takeaways that I have from the session. Nice.

B (6:16)

So I want to come back to this notion of differing futures. And it sounds like you're less invested in exactly how the future is going to unfold versus making sure that you're prepared kind of no matter how it unfolds. You mentioned a notion basically of lesser AI impact, more like today to greater AI impact. One of the other lenses we look at this through often with guests is, you know, kind of positive or negative AI impact, I guess, if you can call it that, if it's going to end up societally building more of a utopia versus, you know, more of a threat. And I guess from a workforce perspective, if it's more harmonious with, you know, people plus AI versus AI instead of people, do either of those lenses, you know, lend themselves to this exercise or do people look at some of those macro factors as they scenario plan?

A (7:12)

Yeah, Jeff. So we have this robust, like what we call drivers of change that look into things that are not directly related to the technology, but all can have a contributing factor in the future that we're engaging with. So a lot of different perspectives can kind of come into this. There could be a negative sentiment towards AI because it has caused either, let's say, the next wave of financial or economic crisis, or because it has caused mass employment, mass unemployment. But I would say, like, specific to your question, in this scenario planning, where we really kind of shift the difference. So there are, I'll back up a little bit. There are four scenarios that we laid out. There's a constraint scenario. This is the scenario where the AI doesn't look fundamentally that much different than it does today. But there's a lot more regulation in place and there's a lot more careful consideration that it's not whether or not we can do this, it's whether or not we're allowed to do this. So that's the constraint scenario. Then we have a growth scenario. This is kind of the scenario that often the headlines that are big across the AI space refer to. And in this instance we have agents starting to run end to end workflows and creating a lot of increased productivity. A single person can now accomplish the work of 10, or a team of 10 can accomplish the work of a hundred. But what we think in this scenario that we lay out is that it's more going to be people directing AI to be able to go do work, right? So still people are in control, they have their agency. The next scenario that we, that we laid out was one that we called a transform scenario. The transform scenario assumes that there is a inflection point which causes the AI and the world that we live in to be kind of a little bit more unrecognizable from where it is now. Now this is more of like the AGI or the, the takeoff scenario. And in that scenario we say that it is more about AI directing us on what to go do. Like it is executing the critical path and, and, and we are accomplishing some of the tasks that the systems are not yet able to achieve. And then finally there's a scenario around collapse. And collapse is not really know, it's not Skynet or Terminator. It's more the scenario that value collapse to a single company or a single country. Because this is a matter in a lot of cases of national security. Many companies are really working hard to be able to compete against one another in this space. There's talks about recurring recursive self improvement. So in this scenario it's like how do you protect and you know, treatment what is distinctly yours in a, in a place where you are renting the AI capability and, and where, and how do you create value in that space? Now the interesting thing is before we do these sessions, we ask people to raise their hand on which they think is most probable. And we've been doing this, we now have like over 2000 votes because we do it at a bunch of different, with a Bunch of different audiences. And it seems to be more like 10% of people are saying Constrain, about 35% are saying growth, about 50%, about 45% are in the transform scenario, and about 5% are in the collapse scenario. So these things are very influenced by the news that people have. But for an organization that has to make bets today, sometimes just looking beyond the incremental and thinking about how you compete in a world where everybody has this very powerful AI system, like the question is no longer should we adopt this or how do we adopt this? It's more a question of how do you win relative to your competitors. And that's the part of the exercise that helps many of the participants start to see that maybe they're not playing to win yet. And there are some other things that they can do to be able to try to, to try to go in a different direction.

B (11:19)

If you work in it, Infotech Research Group is a name you need to know no matter what your needs are. Infotech has you covered. AI strategy, covered. Disaster recovery, covered. Vendor negotiation, covered. Infotech supports you with the best practice research and a team of analysts standing by ready to help you tackle your toughest challenges. Check it out at the link below and don't forget to like and subscribe. I want to come back to the how youw Win in just a second, but just before we do, I want to talk about those stats about the future people are envisioning. And I'm fascinated. That transform is actually the biggest fraction versus, you know, even grow. Has that been growing over time? Like have you seen longitudinally over the past year or two that more people are shifting their vote to transform?

A (12:09)

Absolutely. I did this experiment for the first time a little over a year ago and at that point it was 0 were in the constraint, 50, 50% were in the growth, 35 were in the transform, and 15% were in the collapse scenario. So there is a clear shift of those that are kind of moving their votes in a way some towards the constraint scenario. And that is mostly due to those that are paying attention to some of the economics beyond around some of the data center build out. And a lot more of those are shifting their votes towards the transform scenario.

B (12:53)

It's interesting to see that shift and to your point, how news narratives are shaping that. And I don't want to diminish the impact of the technology itself and how it seems to be improving at such an impressive rate. But I did want to talk about winning, which you just mentioned and what you had said is that winning and the teams that you saw as best positioned to win were the ones that had novel business models, that they were building new business models with AI or in the face of AI. And that's really interesting to me because the other thing that you mentioned that it's not is workflows. Right. It's not just bringing an AI to identify what you're doing today. It's answering some more, as you said, more creative, more foundational question. Can you maybe elaborate a little bit on the differences there and why you see the business model piece as being so fundamental?

A (13:56)

Yeah. So let me back up a little bit. I'm going to answer your question, but let me back up a little bit with just an observation that we've had. You know, about a year ago or a year and a half ago, we stopped talking about AI in PowerPoints and we started talking about AI by getting the opportunity to put your hands on the keyboard and create more immersive experience. So we called these, these masterclasses. And these masterclasses were essentially a way for executives to be able to get their hands on the keyboard. And we created like an impossible task. Once it was to create a shoe and to market and advertise the shoe, and they all, you know, the executives and, you know, it's either a board or the C suite, would sit around the table and start to go create a product, create a brand, market it, figure out how they were going to roll it out. In some cases, we did it around snack products. And I got to say, like, it was a really powerful opportunity and a really powerful exercise for people to put their hands on the keyboard. Sometimes it still surprises me, like the number of, of senior leaders that have not yet gone in and started vibe coding, for instance, just to be able to try it out and get a sense of where the state of the capabilities. But after we did that exercise countless amounts of times, like for, in the, in the snack example, we looked at the results. And while there was incredible passion beyond the snacks that the teams were creating in the room, and you'd hear some giggling. I once had an executive who had a Greek background stand up in front of his leadership team and do a jingle in Greek about why his product was going to be the best. But observing each of the products on a longitudinal basis, like after having done these sessions, many, many times, you start to see that a lot of these products look the same. And it's something that we called the sameness trap. Like in this instance, like all the products had Matcha in them or they had Some like, sweetener that was monk fruit as a way of being able to, you know, a low calorie way of being able to introduce some sweetness into the snack. But they were all the same hues of brown and green, etc. So, so we named that, we started. We call that a sn, a sameness trap. We say that, like, this is where you need to break through the sameness trap if you want to differentiate yourself from other companies. Because everybody is using the same models in a lot of ways. Just using AI commoditizes the output. It takes the creativity and the ideation, the origination, to be able to go build something new. That's what we saw in those winning teams. Like, those winning teams were thinking a lot more before they were just deferring to the AI system in this instance. And they were coming up with creative ideas of how they might be able to not just do what they do today, but kind of reinvent a new, a new proposition or a new product. And this is starting to bear itself out in some of the statistics that we look at. Jeff. So we run a study every six months that we call the AI Pulse Survey and essentially goes out to 500 executives and asks them a series of questions about their investments, where they're seeing return on investment, what are some of their individual, like, their big bets and priorities. And then we ask some topical questions in every round just to be able to kind of get a flavor of what's contemporary at that particular moment. And one of the questions we've asked is where they've seen the most amount of benefit over the last, you know, you know, at that particular time. Now we're on our fourth iteration of this, so we have about two years worth of data. And for the first 18 months, we saw an increasing tick up of companies that had invested over $10 million in seeing their benefit come from employee productivity and operational efficiency. So it kind of gets to that question that I know you've been discussing on the podcast of like, is this something that is really just looking at eating hours of, you know, or going after the jobs that people do today? Interestingly enough, this latest iteration, we saw that adjust in a pretty, you know, in a pretty material way. The, the employee productivity and operational efficiency had gone down as an area where companies were seeing their benefits. And instead competitive differentiation and new product introduction were the areas that had grown the most. Like, I think competitive differentiation grew 10 points in those six months. So I think, I think we're starting to reach an inflection point now where Many organizations are, instead of looking at how they can apply AI inside their organization, are now starting to ask the question, do I have the opportunity to do today something that was yesterday impossible? Like, can I go tackle something entirely new? That was what we saw in the Winning Answers. And I think we're starting to see companies shift their direction of travel in that way.

B (19:05)

That's really exciting to hear. And for me, in some ways, that's like the final frontier of AI, Right? It's not just like, it's not just automating what you're doing. It's having these new frontiers and being able to do something, I think, in your words, that was recently impossible. You intimated that there's a right way to do this and a wrong way to do this, and there's a place that AI can help, and there's a place that AI cannot help. Right. That this sameness trap. Can you give me a little bit more of a flavor of the playbook of, as an executive, what do you really need to get right in this space if AI is going to help you differentiate, differentiate yourself competitively versus, you know, that the initiative is going to fall flat on its face?

A (19:46)

Yeah. Well, I think, you know, first I would say that AI is a tool that really lifts the floor. And what we're seeing is that it's people and their creative thinking that lift the ceiling. Just starting there, I think, is a very helpful framing so companies can maintain their differentiation and. And can maintain what has helped them win up until that point of time as they move forward. But, you know, I'd say that, you know, so you can, you can, you know, ultimately, as. As you've talked about on the podcast before, the goal is to really transform and reinvent or reimagine a new way of working. And there's two ways to be able to do that. You know, like, it's kind of like you're building a plane while you're flying it. At the same time, you need to be able to both go broad across the organization to empower and lift everybody up across the organization and pick a couple of areas where it makes sense to go deep and go through this reinvention or reimagination process to be able to go build something entirely new now recipes to success on those two different axes. If we're focused on the going broad right now, statistically, another. Another statistic that we have is that 88% of people are using AI as part of their work. Like, we don't have an adoption challenge anymore, you know, but what we do have is only 5% of those people have reported that they are using AI to do something different than what they've done in the past. So it's like 88% are using it, 5% are using it in a transformative way. That gap, that 83% gap, is the opportunity that we now have ahead of us that we need to close, and it's not going to close itself. So it takes specific training and development culture and mindset to be able to teach people how to do more than just adopt these tools. And you know, in EY we say that this evolves looking at the sequence. So we need to do our critical thinking and context gathering first before we engage with the AI. It's kind of just to back up. It's human AI and then human and it's a lot of critical thinking, context building, strategic planning. Then we engage in AI in the execution of the work and then finally we have people come back and take a look at the results, start to really challenge the results that we see, make sure there's no biases in the answers and then move forward. That happens in the way they engage in their desktop productivity tools. That also happens in the way that they work in their vibe coding tools. To be able to go build new applications, to be able to power the business, there needs to be a lot of critical thinking up front. And often right now, the training that enterprises are putting out are more tool focused training on how to use these particular tools. So we say at ey like there's it takes three sets, takes the right mindset, the right skill set and the right tool set. And unfortunately, most of the money today is going towards the tool set. And that actually needs to flip where the money needs to be going more towards the mindset and the skill set while the tool set is in place at a lot of organizations. So that's kind of a playbook for how organizations can teach their people not just to use AI, but to use AI in a way to be able to create or drive something that was previously unattainable or impossible at the point of time.

B (23:27)

It makes complete sense to me. And one of the questions that I've been grappling with, and I heard it come out there a couple of different ways, is to what degree success in getting the most out of these new business models and new tools has to be top down versus can be bottom up. And the reason I ask that is because it sounds like on that ending note there, Dan, there was an implication that if you train people not Just on toolset, but on mindset that maybe more of this can be bottom up and you can yield some of those results. Do you buy that or do you think this really needs to be driven as a top down exercise?

A (24:11)

You know, I, I know this is still something that's being sorted at the moment and many organizations are on this journey. I still think we are in the early innings of a multi year business transformation and it's a lot less about the technology, it's more about changing mindsets, behaviors and institutional ways of working. But I kind of, with respect to the going broad and going deep, I do think that needs to be led top down inside the organization. Because when you start to do things in bigger and bolder ways, you often tackle things that were not clearly in the lines of the existing organizational chart. You know, like the organizational chart has been built around, you know, old ways of working and when you start to apply AI, you start to stripe across these different, these different org charts and in doing so then you start to need support from the top to be able to really push agenda broader. However, that is not to say that it's, that's all that it takes. I mean I, you know, one of the other things that I think is, is, you know, an observation that we have in working with many of our clients is sometimes the deepest expertise that you have actually blinds you from what the new process can be. Something, you know, I, we've called these things traps. So we call that the visibility trap. That like you know this process so particularly well that you can't imagine a new way of working it. It kind of blinds you and locks you in. So the, so the antidote to that actually has to do a lot with the bottoms up. You know, sometimes we find that the best way to be able to go reinvent the work is not by having the domain experts take the lead and have them be supported by AI people. And it's also not by having AI people that know nothing about the process lead and the domain team support it. But it's have an equal footing in the design between those two that you bring your domain experts together with people that know nothing about the process but are more native in the way they think about AI. And Jeff, many times this is people that have come right out of school. You know, it's kind of this interesting phenomenon where I know many, many students that are graduating soon are looking at the prospects of a harder job market. Sometimes those people that have the ability to say why do we do it this old way? Like you know, this does not make sense anymore are the way that that innovation can happen. Bottoms up. So. So while I do think it needs to be the conditions for change need to be set tops down, it does not mean that the initiatives need to be led top down. You can start to leverage that talent that exists that is much more fluent in working with these tools to be able to challenge the status quo.

B (27:05)

I haven't heard it articulated like that before and I really like that sort of hybrid approach where you can have people within the organization with either domain expertise or more as generalists and don't necessarily have the baggage of the way we've always done things. That, that, that makes a lot of sense to me.

A (27:22)

I'm curious if you, if, if you have a problem, if you're a leader and you have a problem that you have to go solve, the normal instinct is to put the person that knows the most about the problem to go solve it. Like that is totally normal. But, but sometimes that can just lead you down a path that you solve a tiny part of the problem. But you don't really challenge the way that you know the, the entire process from an end to end product perspective.

B (27:46)

Wow. And you have a consulting background, so I'm sure you can appreciate that the more you do that, the more likely you are to end up at a target state that looks an awful lot like your current state. Right?

A (27:54)

Yeah, yeah, yeah. In a lot of ways, when you have to reinvent, you almost need to forget the current state. You know, we say like break down the jobs to be done or the key decisions that need to be made to be able to go run this particular function. You know, sometimes we refer to it as zero based design or we have, you know, kind of a reinvention methodology and that helps you start to really, really unencumber yourself from the institutional legacy.

B (28:22)

So I'm curious when it comes to AI and it comes to business model transformation or operating model transformation, historically with these types of transformations, it seems like there's some degree of buy in from everybody that yes, it's going to look different, but we'll all kind of chip in and make this work. And there is a decent, as I'm sure you've seen the stats, there is a decent failure rate. I'm curious to what degree you're finding with AI that AI's reputation and some of the stories you hear in the news about AI just taking jobs or making people obsolete, that some of this fear and uncertainty around AI is preventing cooperation at an organizational level because people are so Concerned that AI is going to take their jobs. Are you seeing that play out in practice or is that more of kind of a, a headline problem?

A (29:17)

I would say that there's always resistance to new things. You know, people as, as, you know, on a continuum, about 10 people like change. You know, on a continuum of a hundred people, ten people like change and about ninety people really do not enjoy having, you know, the proverbial car, the carpet pulled out from underneath them on a regular basis. So I think this is something that is, that is actually quite normal and expected. I think that can be very amplified, though, by the way that AI is taken inside the organization. So if we, if we start looking to apply AI by, by taking the process as it exists today and asking the question like, where can AI go do some of this work? You know, we're starting this process of, of an employee feeling like they are facing death by a thousand papers, cuts, you know, because a, a job is something that is artificially made. It's a collection of tasks that people do. We bundle those tasks together and we call it a job. Most cases, AI is not automating the job. AI is automating a bunch of tasks inside the job. And eventually, if you keep going, then you start to feel like there's, there's really nothing left. So, so I do see some resent, some resistance if you take that approach because you're applying AI to a new way of working, but you're not showing where the employees come out benefiting this on the other end. However, there is, there are some different ways to be able to look at this. You know, if, if you take that reinvention or the, what I was calling earlier, I don't want to call it reinvention, Jeff. So if we take that reimagination methodology that we had referred to earlier, then you can start to cast where the jobs, the new jobs are going to exist inside the organization. So instead of saying, okay, well, you know, we're going to automate 20% of your work this week and the next week, we're going to start looking at where to go next. Instead, we start transitioning people from being a domain expert that is responsible for conducting this work into a job of a knowledge engineer where they are responsible for collecting the right intelligence and the right context for these systems to be able to work in an optimal way. And we're going to push people more, or allow people to go more into some of these creative spaces where they get to sit atop and start to orchestrate these agents on their behalf. That story of what happens next is often missing when you start to run into the organizational resistance, because you're just looking for where AI can solve a problem and you're not showing where and how the employees are able to unlock new possibilities on the outside and where they might land into a job. That is absolutely going to be required when we fully implement AI inside the organization. So those are just some ways that we've seen organizations be able to get through navigating this. But, you know, this is pretty common. Like anytime you go through a big period of technological revolution, there's always a lot of concern about disruption. And then usually if the past has any indication of the future, then usually you end up reinventing a whole bunch of new jobs that come on the back end of us.

B (32:41)

I'm curious to that point, as you think about those new jobs, what they are, and what the future of work looks like. There was a conversation earlier about kind of mindset, skill set, tool set, in that order. Is that going to hold what are the skills of the future and what's most important for people to be good at or what's going to help them bring the most value to the jobs of tomorrow?

A (33:08)

Yeah, as we kind of had referred to earlier, if we work with AI but don't give it the appropriate context, the appropriate amount of direction and creativity, if we don't start to give it the counterfactuals to be able to make sure that it is operating and giving really useful responses and content, then we're going to go down the path of statistical sameness. So we see that there are a couple of really important skills for the first future. So there's a skill around critical thinking which is just, you know, avoiding a lot of these traps where, where, where you can fall victim to an AI that is behaving either in a sycophantic way or is really aligning you down a path of answers that would probably be applicable for a whole breadth of questions. There's a skill around, around deep domain expertise, and this is where you can provide the appropriate instructions for the systems. There's a skill around management and delegation, and this is something that's really important in the future. When we hire people into, into ey, we are now going to start training them more on managerial tasks like the corporate ladder used to be that you worked, you got experience, you did your time, and then if you were lucky, you got a promotion. And in a lot of instances, now we're expecting people to manage on day one. They're just going to be managing a bunch of agents that, you know, are clumsy but very powerful and can go do a lot of the work that they would have traditionally done in the past. And I'd like to come back to a point around, like, what that means for the apprenticeship model, but we can do that a little bit later. I want to finish up the skills here. So critical thinking, the creativity, the, the, the domain expertise, the managerial skills, and then big picture thinking is actually really important. Like sometimes people refer to this as systems thinking. Being able to kind of see the big picture and zoom out. These are incredibly important skills that workers will have in the future because they will move from doing the work to orchestrating a whole bunch of AI doing the work for them. And, you know, you're only in that case going to be as good as the idea and direction that you gave at the outset.

B (35:32)

Can we talk a little bit about the role of IT or enterprise technology as a function within the organization with success here, with bringing these tools in, with training the business function staff on this? What does this look like? And I guess the backdrop of this question is there's an awful lot of obituaries being written right now for business functions for industries that AI just none of these matter anymore. And one of the trends I've personally seen around AI is this rush to business adoption where it's almost happening around any sort of corporate it. And I'm curious what role you see a core technology group is playing, you know, how important it is and what the risks are of not having that at all.

A (36:28)

Yeah, great question, Jeff. Yeah, AI is an incredibly democratizing capability that since everybody has it on their phones now, everybody brings it into work. As I mentioned, in the recipe for success, there's a combination of going broad and going deep. And in the going broad aspect there, the real role of technology is to be able to empower the workers with as close to the frontier capability as they can, but with the appropriate guardrails and sandbox. To be able to do this in a way that respects information security, privacy, intellectual property, you know, all the, all the ethics and the corporate usage standards that exist. So that is much more of an enablement. You know, maybe you might argue like a traditional technology function where we are providing a capability and scaling that across the enterprise. But when you move the focus to going deep in a particular space, the role of the technology department is going to become more of the execution of the operating system for the future. Because the real hard problems to solve are two. It's how do we orchestrate These workforce of agents to be able to go do the work with the appropriate security and access controls and closing off this new larger surface area for, for cyber risk and for a bunch of, a bunch of, you know, infiltrations, let's say insider risks. So that's, that's one is that orchestration layer is essentially the new operating system for many of these companies. And the second that really help you get the most value out of these agentic run operations are maintaining the kind of being the library of the corporate memory for the company. Because context is what powers these agents to be able to do work that works for you, Jeff, or that works for me. And the same agent can work in a completely different way for you than it will with me if we configure it the right way with the right appropriate context. But most technologists and most technology does not address this corporate memory in an agent world. So this is going to be an area where I would expect many of the technology departments inside of traditional companies going to need to move towards and figure out how they do that while also maintaining and managing the legacy landscape, the legacy software landscape that they have. I mean, many of the conversations that we're having with technology leaders now is what do I do with my SaaS products or my biggest, my big platforms that I've used and have been the center of my company for the last couple of years and how do they coexist with this world of agents? So I think they're going to have a very important role in defining the operating system for the future of the company. And in that respect it's kind of a two speed job. Enable everybody with a capability as close to the frontier as you can and then in parallel start to define what that operating runtime system is in the future.

B (40:05)

I want to push you a little bit more on the question that you were asking broadly about how these systems do coexist. When you've got your enterprise CRM or ERP system and maybe you've been working on some sort of legacy migration for the past nine years or something. But again, one of the obituaries we hear about these days is like the SaaS industry in general and you know, who needs SaaS when everybody's, you know, coding their own apps and Claude, you know, in their spare time off the side of their desk, what does that interplay look like? Do you see the future as totally moving off of these platforms? Do you see an integration layer? Is it, you know, two completely different speeds and lanes and functions? What does success look like? There.

A (40:52)

Yeah. Well, for the last 20 years, many companies have been told that custom development is too expensive. So therefore, the answer is to adopt the standard and fit your company into the standard that exists, because that is going to be the most efficient to run, that will be the safest and most predictable. And this is the standard. And that has worked for a lot of companies. But what has happened is that standard has started to sprawl out from what its original core is. And what you have is, as it sprawls out, you have to spend a lot of time maintaining the custom aspects that you've put around that particular standard. So what we often refer to inside of EY is we say there's been this buildup of a sprawling core and there's a starved edge. Like we've all probably, we all know of companies that have billion dollar ERP programs, but when you engage with their sales force, like the systems of the systems of their customer engagement, and I don't mean the product salesforce, when you engage with their sales teams or when they engage with their customers, everything is done in Excel. So in a lot of ways, there's been a buildup at the center and neglect at the core. And for the, for the next couple of years, I would expect that to basically, with these new tools to be able to address this starved core with the kind of new capability that will allow them to be able to be a center for differentiation in the future. So I don't think these things collide. You know, anybody that's saying there's going to be the obituary for the CRM or the ERP at these large companies, you know, in large Fortune 500 companies. I think we're still a little early to start to draw some of those conclusions because it's not just the application. Maybe you don't like to use the application, but it's also all the authorization objects and like all the, the privacy controls that are built into those systems. But you can build the, you know, custom software to be able to address the parts of the business that have not gotten any capital, that have been starved for investment over the course of the last decade and start to see a lot of real build out in those spaces.

B (43:15)

That, that picture you paint is really interesting. And I like the metaphor of kind of the core and the edge there. And if I'm understanding correctly, Dan, you know, one of the things that I'm concerned about is when you look at, you know, a particular salesforce or team that needs to use these tools and is not, you know, is Getting starved for the tech that they need. In this world where there's a commoditization and consumerization of AI tools, maybe they will build their own apps and maybe they will be able to, you know, have all these, these frontier tools that don't necessarily play nice with the corporate core. But your point is, well, how is that, how is that worse than just using their own Excel anyway? And that it's a problem, but it's just kind of a continuation of an existing problem versus, you know, an entirely net new problem. Is that fair or would you, you know, conceptualize that a bit differently?

A (44:10)

Yeah, I think, I think the general thesis is very fair, Jeff. But just something I'd like to, to clarify is, you know, I wouldn't expect that sales reps would start to build their own applications because they have access to these tools. I would expect that you'd have a couple of sales members working with a couple of AI powered software engineers, like people that really know how software can be built to design a system that is able to solve a problem that exists in the salesforce and stand that up in weeks where otherwise it would have taken months or years to be able to make happen. I think the build your own, the build your own application is really interesting for prototyping and for ideation, but I think that a lot of the value that happens inside of these applications is the concentration of insights, the concentration of proprietary data, the network effect of, by people using this application. Like the more I use the application, Jeff, you and I work at the same company. The more you benefit from that. Those start to really compound like knowledge and data start to compound over time when there are, when there are applications and products in place. So I wouldn't expect it to be a thousand flowers of, of products across the organization, maybe a bunch of different proof of concepts or prototypes that would exist and that all to feed into the opportunity for a company to now invest into a space that previously they just never had the capital to go to.

B (45:58)

Right. And that's, you know, I'm not going to drain that one because it's quite aligned with my thinking as well. I mean, that's exactly my view. That great for pilots for proof of concepts, but as soon as something is, you know, has proven itself and you want to do it at scale, then now it's time to actually figure out how you integrate it into the core. Yeah.

A (46:18)

And that is being AI powered as well. Like in many instances when we use some of these vibe coding tools, we use them for the purposes of being able to derive a working prototype. And, you know, in the future, I anticipate that we'll be able to treat prototypes in a similar way that marketers think about different pieces of, of marketing content. And we can do AB tests on what is the most effective to work, but then you have to put them through an enterprise standard. You know, this is, while this is an incredibly powerful capability, it also expands the surface area pretty rapidly of where, where there are the opportunities for risks and intrusions. So, so then you consolidate, you build a standard that can be powered by AI as well. So that can still. It doesn't mean that we got a prototype in a day and now we need to go do the project in 12 months. That can still happen very, you know, very fast. But I think that is a job that will likely done by people that are professionally trained in that, in that space, as opposed to the, the people that are building out the prototype.

B (47:18)

It makes sense. So, so just on the topic of people who are, you know, leaving it to the professionals, let's say I wanted to pivot to a slightly different area. And you know, Dan, I'll preface this one by saying I know I'm not asking a neutral party here, but one of the obituaries you hear about from time to time is the obituary of the consulting industry. And why do you need consultants when you can just ask AI what to do? And so I'm curious, and I'm being, you know, semi facetious with that, even though, you know, there's lots of people out there who, I'm sure, you know, are completely convinced that it's a, that it's a debt industry. How does consulting reinvent itself as an industry and as a value driver in the age of AI? And what areas of consulting, if any, do you see as being under threat and maybe need to die to make way for this new age?

A (48:11)

Yeah, well, thank you for asking the question, Jeff. I would say that a little bit earlier in the conversation we had talked about the fact that in order to get more out of AI, you need a couple of core skills. You need to be able to do the critical thinking, the creativity, the management and delegation, the systems thinking, and have deep domain expertise. And when you can put those together, you can start to create things that are differentiated and things that the AI would have intuitively known to be able to go build. When you look at those five skills, those are the skills of most of the consultants that exist today. Like those are core competencies for consultants. So, so while I think in a lot of instances the Work the assembly that people in the consulting industry do will change. Like, we're not going to spend weeks working on PowerPoint. It just allows us to be able to close the loop in terms of the speed in which we can work with our clients to be able to create new areas of opportunity and growth for them. In a lot of ways, many of the traditional projects will have changed and some of them will go away and some of them, I think some of the ways that we do our work will just be significantly enhanced, but I think it will open up a new opportunity space for us to be able to help shift focus towards growth and growth without the budgetary constraints that many of our clients have had to be able to go examine and evaluate possibilities of ideas. So for that reason, I mean, we hire extremely talented people that have a lot of the core skills and requisites that are, are good for getting the most out of AI. And for that reason, I wouldn't write the book on the death of consulting yet.

B (50:13)

It makes sense. And just as you were explaining kind of your view on that, I realized that I didn't ask something earlier that has been kind of piquing my interest, which is when you describe the four futures and people voting on which future they expect to see. What, what's your vote on that, Dan?

A (50:33)

Oh, oh, this is, this is like different because in a room I can just, you know, I can just say it comfortably, you know, I've spent so much time. Jeff, I'm gonna try to not give you the non answer, but like, I can, I can convince myself of any of these things happening. And that's why, that's why they're real and plausible. It, it shifts, it shifts, you know, quite quickly. You see like all the, you know, the gigawatt data centers that are, you know, are projected into the future. And I wanted to really get an appreciation of where the bottlenecks exist, because there have to be bottlenecks. Like, we can't manufacture that many new chips at the kind of speed that people are referring to. So I went deep into learning about how the ASML machines work, the EUV machines, which is like, it literally feels like science fiction in reality at this point in time. And if you like to, we can, we can geek out on that a little bit. But that just helped me kind of see that there are some constraints that we have, you know, coming down the path, you know, that in a lot of ways there's, there's a lot being built up that we hope, you know, will kind of see its way through and there's new things that are affecting the world now, you know, wars, et cetera, that can put a lot of pressure on inflation and those sorts of things. So, so I guess I'm basically giving you the preface of why I probably lean a little bit more constraint right now than, than any of the other four futures, but that, that kind of changes and I think part of the whole purpose is to, is to start to evaluate your priors and see how things change on a continuous basis. But since you asked me the question, Jeff, now I have to turn the, turn the mic over to you. What, what, what would you say based on the little that I've given you about those futures, what, what would you lean towards?

B (52:35)

Well, just before I answer that question, I was going to say I find it very interesting that you lean towards constraint given that that's the, the area that seems to be less popular. If I remember your numbers correctly. I think it was like the third place answer. Yeah, I'm tempted to answer somewhere between growth and constraint and I know I'm deliberately picking somewhere right in the middle. But I see what's going on with constraints of this technology. I see the unevenness of the adoption and it's difficult for me to jump right into the transform future at least immediately. If we're talking about 2040 or beyond, I can get there a little bit more easily. But in the next few years. Yeah, I can see that happening very quickly. And it's interesting you bring up the AI infrastructure piece because one of the things that's been catching my attention lately is with this sort of gold rush in AI and a lot of the technologies that you need to build this, whether it's rare earth minerals or whether it's ram, how it's changing the economics of adjacent industries and how that's leading to sort of downstream constraints and none of this is just all net positive, zero impact on anything else and how these trade offs are going to, you know, in some ways destabilize, you know, the economic model of today.

A (54:07)

Yeah, absolutely. I would, I. Since you chose two answers, I would say I'm right between those two of constraint and growth. I, I picked one. Follow the instructions, but thank you. But, but it's, it's right there on the border. For those reasons, like I think the technology is going to continue to, to take off and going to continue to grow. I'm not, I'm not suspecting that the tech like the AI that we see is as good as it is today. I just wonder, you know, can we can we keep the token price, you know, as what it is with, with demand, you know, continuing to increase at what it does and with complexities, you know, you kind of referred to it, but almost all the semiconductors in the world are, you know, are made in a single country that's, you know, located on a fault line and is an area of geopolitical political tension. So yeah, there's a lot of. It's, it's amazing the more you spend time looking at the, the, the semiconductor industry and the supply chain for AI infrastructure, it's like, it's fascinating how it all works because there are like multiple single points of failure along the entire chain that, that are all working in concert at the moment.

B (55:16)

Yeah, yeah. And that, that very naturally takes you right to that constraint concern. I am curious though, and one of the things that's interesting about being the host on a podcast like this is I get to hear all sorts of different perspectives around AI and the future of technology. And as you might expect, most people's perspectives tend to look an awful lot like, you know, whatever product or service they're selling or however their income stream is comprised. So you end up, especially with the big tech companies, they tend to be selling you this vision of transformation versus this vision of, you know, collapse or market capture, which I think there's probably more truth to than they would like to admit, with narratives that are not necessarily neutral, just taking up so much of people's mental bandwidth. I'm curious from your perspective, Dan, if there's any narratives you're hearing around, you know, AI or future of tech or future of work that you think are bs, that you hear a lot about and you're saying, you know what, whatever you hear about that, that is somebody pushing an agenda or it's a compelling, you know, attention point. But I don't think there's a lot of merit to it.

A (56:29)

Yeah, Jeff, I think, I think humility is important in this space, to be honest. So anybody tells you that they know exactly where things are going to end up with, with like a high degree of certainty, unless if they're working in the labs, like in the labs that the hundred people that are working on the product, then I think it's a lot of speculation at this moment. There's still a lot that's unsettled. I think there's a, there's often a bit of a narrative that, you know, people will be relegated. I think in some industries we're already starting to see what happens when you over automate and just because you can automate think, something doesn't mean that you should. That like, ultimately, at the end of the day, value is created by meeting a consumer need, and consumers ultimately have the buying power over where they go. We've already started to see some instances of companies deciding that experience matters a whole lot in what they buy. And that experience is different if you have the experience with a person than if you have the experience with the company's AI robot. So I think the narrative that often is that people will be relegated to the side is one that kind of fits in this cone of possibilities. But I think at the end of the day, given the fact of who has the buying power in the consumer space, I think it is. I think that is significantly overplayed right now. And I think it really devalues the real power that humans and people bring in being able to create new ideas. So far there's been talk about this, but a lot of the new idea generation is still a very, very human thing, and I think that's underrated at the moment.

B (58:31)

No, I appreciate that. And there was an insight in that answer that I really liked, which was that it seems like enterprises are maybe over indexing on automating what's easy to automate versus what's high value to automate. And it could be actually, to your point, automating the wrong things just because they can, but they don't know if they should. And that enterprises can be complicit in that. And that's probably not the road forward as we look at the future we want to build.

A (59:07)

Yeah, exactly. And diffusion, the term that's often used to be able to say how long it takes for AI, the state of the AR capability to be able to asseminate across the organization like that is, still takes time, you know, so sometimes the, the, the projections of how deep or how impactful AI is going to be in the short term seem to be, you know, missing the part that at the end of the day, this is all about change. There's a famous quote, you've probably heard of it before, Amara's Law. And if you don't know what a Meyer's Law is, you definitely have heard of what the Law is. But it basically says we tend to overestimate in the short term and underestimate in the long term. And I think many that are in the news talking about AI are kind of falling into that trap.

B (60:09)

Right, that makes sense to me. So, as we start to wrap things up here, Dan, I'm curious if you have any Parting wisdom you wanted to leave for business leaders or technology leaders in terms of your top advice around what they should be thinking about or what they should be planning to get right to make sure that they are able to future proof in this coming age.

A (60:32)

That's a tough question. At the end of this dialogue, I'm trying to think about what I haven't shared, but I, um. At the end of the day, you know, senior leaders have a really important role in shaping the direction of where the strategy goes across the organization and how much they are doing this with their people as opposed to against their people. So you know, I think where I've seen organizations really lean in and leverage the collective wisdom and brainpower from their organization is when they move more into a value creation mindset as opposed to an operational efficiency and a cost cutting mindset. When you go into a cost cutting mindset, you're turning off the part of the brain that really focuses on creativity. When people feel threatened by a technology and move to a state of fear, they're starting to cut off the part of their brain that focuses on creativity. So in a lot of ways, the way that AI is discussed across the organization as lifting the floor so people can now lift the ceiling is incredibly important to bringing the organization along. And sometimes in this space of AI we look for trying to drive adoption and we look to find where the value is created and we fall down that visibility trap of looking to where the. We apply AI to the work that we do today and just like finding a bunch of incremental use cases that don't really substantially move the company forward, but start to start to do that what we referred to earlier as death by a thousand paper cuts. So I think positioning and strategy is so important for organizations and it's often left to the team that is responsible for drive in the AI program. So that would be just one part of spend time as a leadership team getting a common understanding of what it is and how we're going to use AI to be able to, to to to power the business. What are the new markets we can move towards, what are the ways we can enhance our products, what are the new opportunities for us to be able to grow and if the executive teams can stay consistent and locked in with that language and that will be a very inspired organization to be able to go run through walls for you. Everybody has an understanding of what AI is. I don't need to do learning sessions on this anymore. But everybody's definition is what it means for the organization if they haven't done. That session is a little bit different. And if that happens, that's when there can be fractures and there can be organizational resistance to change.

B (63:13)

I love that and I think a really valuable note to end on. Dan, I want to say a big thank you for joining today. This has been a really interesting and insightful session.

A (63:22)

Thank you so much for the opportunity. Jeff

B (63:26)

if you work in it, Infotech Research Group is a name you need to know no matter what your needs are. Infotech has you covered AI strategy, covered disaster recovery, covered vendor negotiation, covered. Infotech supports you with the best practice research and a team of analysts standing by ready to help you tackle your toughest challenges. Check it out at the link below. And don't forget to like and subscribe.