Transcript
Evan Metter: Hello, this is Evan Metter and I am so pleased to bring you Keeping Peoples Minds Growing a podcast that was inspired by my daughter asking me one day what KPMG stood for. I told her about the origins and she came up with something better. On this podcast we are delving into provocative and insightful conversations that are not only fun to listen to but seek truth and perspectives from those who are at the forefront of change happening in the world of work. So if you have guest suggestions or topics that you think would be fun to cover, please email me Evan Metter at emetter@kpmg.com. And now, let’s jump in.
I'm Evan Metter. I am really pleased to be hosting here with my good friend Mike DiClaudio on today's inaugural episode. Our vision for this was really to create a space for real dialogue about the talent and human capital issues that we're experiencing not only as we work with organizations around the world, but with our own organization. We're all going through a massive disruption and change, and it seems like an open time to have more open conversations, so I'll just get it kind of kicked right off, you know, Mike, one of the things that I wanted today's conversation to be about is AI is, is they can't go anywhere without talking about GenAI. I've been thinking a lot about how that actually affects talent, but really entry level talent and the conversation around, is it gonna essentially obviate the need for entry level talent and then what is the talent pipeline and I just start out like what's your hypothesis?
Mike DiClaudio: Hi I'm Mike, I'm the head of talent for our advisory business. The general consensus in the market is it's going to kind of raise the floor of the organization up. So we're going to be doing away with entry level jobs because I can automate all the tasks that I used to send to entry level workers or I'm going to need fewer entry level workers because those jobs are going to be subject to mass consolidation because a bunch of the tasks inside of those jobs can be automated, so I need fewer hands to produce the same amount of output.
Evan Metter: So, you have kids that will be in the workforce in what 15 years, give or take. Get out of here. So similarly, let me just give you, here's what I'm struggling with. I agree that every time some new technology comes, it's, you know, doomsday. There will be no jobs that will be created. Is there, is there something different going on? So, like yes, I think there's a short-term gap in terms of like there's data our team has pulled that 1 in 12 executives have stopped hiring for entry level roles, and whether you believe this or not, you know, it's in the zeitgeist. You have got 86% of executives plan to replace entry level roles with AI, you'd be foolish not to. So, I think we might be talking about different horizons. Like if I was an entry level person now versus the longer term, and I can, I can appreciate that. However, here's what I'm kind of got my head wrapped around and I'm worried about Mike, which is you and I came up in a world where we wrote things out and I go back to this quote where I was talking to one of my early mentors and he was talking about how he hung this sign this kid's bedrooms being like, writing, writing is not the practice of putting down on paper what you think. It's, it's the practice of figuring out what you think, right? And that idea of having to write. I worry about the loss of that ability to do the critical thinking, because the answers are easy, people will be drawn to the easy answers and the idea of actually thinking through things through creating yourself.
Mike DiClaudio: Yeah, so I think we're talking about a couple of things and putting them all in the same basket right now. So, the first construct around like constructive critical thinking skills, like I like how you phrase it like this is my ability to figure out what I think as opposed to just communicate what I think, uh, that all we're suggesting is I'm giving you a different pencil now. So, the pencil you used to have was a direct line between your brain and your hand and your thinking, and maybe it was going to be informed by the things you had to hand, right? So, it's going to be informed by something in the library, it's going to be informed by family or friends or conditions. Like we always had biases, right? What we're really doing with generative AI is we're exponentially increasing the source of your bias. Like so instead of it being a couple of books in the library, it's the library.
Instead of it being the experience of one or two people, it's the experience of everyone who's ever communicated their experience, and all of those things get funneled. So what you're, what I would actually flip it around and say there's going to be a ton more intuition, and this gets to your second point around my understanding of media bias, but there's going to be a ton more intuition and really the conversation we need to be having with uh with like this is the conversation I have with my kids is like let's learn empathy.
Because really empathy is the key to kind of decoding these things and putting them into your own thought process as opposed to turning into a parrot. Like part of my concern with traditional education systems is they reward like a rally. They reward your ability to kind of rebound back exactly what you're given. And I think what this is doing is dramatically accelerating critical thinking. I don't think it's slowing it. I think it's accelerating it if we do it right, if we allow people to kind of stop at the first answer, like, tell me who is the best baseball team in 2015. Well, that's actually a subjective statement. You can tell who had the best record, but like who's the best team? You're going to get into all sorts of fun debates and this is like a pardon the interruption thing versus just tell me a fact about who had the best. Who had the best record, and if, if you teach, and this is gonna like, if you teach your kid to ask the follow-up questions, well, why was that the best team? How did you think about the best team? That's going to teach him how he should be thinking about the best team, and then he's going to be that fun guy at the bar who's going to debate ERA. So I think what you really want is to be using this as the tool that it is, which is again, it's a dramatically larger data set for people to be reacting to, but make no mistake, people have always been reacting to something and they've always been just playing back what they've heard. It's just that we were an education system that sort of allowed for that. What you need to teach now is the follow on, the follow on, the follow on. Like I need to, I need to start teaching people to be like really good newscasters. Like I need you to interview really well because you're gonna be interviewing a robot and you're gonna be understanding from a robot why the robot thinks what the robot thinks, and then that's gonna help you decide what you think.
Evan Metter: I think that I think you're putting a lot on an individual's drive, right? And I, I think what I wonder, I don't know, what I wonder is, you know, if you were if you were graduating from college right now and you're using AI and, and you can get a lot more done. We were using AI at work and I'm finding that, you know, summarizing emails like it's a huge benefit and it accelerates. I don't know if that has necessarily made me more creative, but it's definitely made me more effective and efficient. And I guess there's two ideas that I've been kind of nurturing and kicking around. One is the idea of wisdom, I feel like we're not old yet. I still, I'm still holding on to the idea that we're, right? We're not old yet, but we have been in the workforce long enough to accumulate some amount of wisdom from like being in the rooms and modeling and seeing, but, but not really, but like doing more menial work to order to be in those rooms, like doing more tactical stuff to be to be able to absorb that. And I wonder one, are the younger elements of the workforce going to be able to be in those rooms and accumulate that wisdom, and or our folks that have come out of college with more experience and more wisdom, because I guess what I'm, what I'm kind of circling around is like any tool, if you came out of college when Excel was new and you had used Excel a lot, you'd have an advantage, right? And I think the advantage still with AI, GenAI is, if you're not using this in high school and college, you're doing yourself a disservice. You have to, you have to do it. But to your point, which maybe is a nuance. Just getting an answer is different than using it to get the best answer, and that is maybe where we're gonna see a different way to evaluate talent. It's like, you know how to use it? Sure, but do you know how to use it to actually get 4 levels deep versus just giving me the top of stack answer, which is what I'm seeing most right now.
Mike DeClaudio: My sister-in-law is a professor of statistics, and one of my best friends is a professor of literature, and they're both having this fascinating reckoning with GenAI. And like so when GPT dropped two years ago, give or take, um, the initial reaction from both of them was, well, there goes the neighborhood. So, there was this huge concern about all the kids are going to do is type in a prompt and then suddenly their statistics final or their initial poetry read is going to all be provided for them. And they had like an initial wave of students that were just not doing the work, like for sure, right? So, but what happens? Like the system adjusted. So they went from saying Ge AI is like the thing I no one's allowed to use it, right, which is, by the way, I think that's like, you know, abstinence training isn't the answer. Like you can't, you can't put that genie back in the bottle, like it's not gonna happen. So, they kind of fairly and they're at two very large prestigious universities and so they very rapidly as a faculty came to terms with like, look man, GenAI is going to be used. What we need to figure out is how we use it.
So, and then how does the fact that everyone in the loop, so we talk about human in the loop all the time, like, how does everyone in the loop use this as a tool both to help inform what you're going to be thinking and drafting and writing as your own, but also how to assess what you're thinking in drafting and writing as your own. So, there's been a really fun evolution that I've sort of watched from two very opposite ends of the academic spectrum, right, like behavioral statistics and fine language arts, and they've both gone through this wave of like, I'm not gonna use it to, I begrudgingly use it to, of course I use it. And it's changed how I think about my curriculum. It's changed how I think about what I asked my students to do. And pretty soon, I'm not gonna have a student that's going to be learning this. I'm only gonna have students that know. If you really want to look for the best horizon point for this, go back to the printing press.
Like what we were really saying was knowledge was captured in very small places and available to very few people. And now I create a model where I could make it distributable to the masses. And I had biased problems because only certain people could afford printing presses, which is, so there was Bibles out before there was like instructions on how to plant crops. But I can go through that evolutionary cycle and say wow, massive expansion to access of knowledge, generally speaking is a good thing. And yet, we all have some turbulence as we navigate that. But getting to that next atmospheric level of everyone having access to much more information has proven to be nothing but a good thing for us as a society. So, what we have to embrace it is the punch line. Like if we run from it or if we view it as a portent of doom, I think we're missing the opportunity.
Yeah, I think so that's right. Um, you would ask a question before I made a statement before around like, you know, it's some version of us what we value, or are we going to change what we value, right? So and I think there's um I got asked this question at a forum years ago about like, hey, where are you going to send your kids to college, knowing what you know about what's happening with disruption? I was like, well, first of all I don't know anything nobody else knows. But the answer to me, and this sounds self-serving because I have a liberal arts degree is liberal arts, like, so what the there are absolutely places and spaces for specialized education that is focused in special ways. The answer is to constantly being able to use those skills in a variety of new contexts. So you talk about like organizations versus individuals like organizations by definition operate at the smallest and slowest common denominator.
Like that's just the way organizations are. Very few of them are agile if they're also scaled, right? So, you can talk a lot about, you know, what it takes to feel like you're more of an entrepreneurial spirit, an entrepreneurial organization, but like if you got like 1 and even 2 commas in the number of your employees, I would temper those expectations. I think there's very much, right, like organizations are built around a cost expectation around shareholder return around risk reward paradigms. So, to some degree they're always going to be um they're gonna be moving at their own pace, let's put it that way. So, which is fine, which is great. The entire economy apply relies on that. I think this question about individuals is different. So, this is like why macroeconomics and microeconomics are different tracks in economics. It's like microeconomics, what you wanna do is to figure out how your skill set, your capabilities, the things you're passionate about, maintain relevance as capabilities and, and technologies evolve. That's your task, right? So that's like, that's the question of like what are you teaching your kids? Well, they're, you know, 12 and 9, so the right right now the answer is complicated, but as 12 and 9 year olds, again, we're back to empathy and we're really focused on learning agility. Like the thing you're learning today, I don't care about your ability to be a rally ball and bounce back. Like I care much more about like, do you actually understand what you just read? And like, what if it had to be something else? And what if you got like a different point of view about that? How would you rationalize those things? Like, um, it's more about like navigating knowledge than expressing knowledge. And I think that's really what we're gonna be looking at. Like there's gonna be this near term disruption because there always is, right? There's always job loss or job movement, like that's always the truth, but I think the winners in this thing are people that can find ways to navigate knowledge versus just being really good at expressing it, and that is, that's a distinction that I think is worth.
Evan Metter: I like that, and I'm going to invite in our producers now, so Phoebe and Julia and Louie grew up in a different period of time, different point in career. I thought it would be useful if the three of you, like, what resonates? What are we not thinking about? Get us some, some, some diverse opinions.
Phoebe: How are educational institutions actually adapting to that? And then what does that mean for how you're looking to educational institutions from, for example, a talent acquisition perspective to actually deliver new students to be future ready for the workforce.
Mike DiClaudio: The upshot of that is on the educational side so like look like our organization like many, has been connected to certain institutions for like a really long time, right? So, like if you go to a school that has a big accounting program, I can almost guarantee that we are connected to the faculty of that program. We know the chair of the accounting department, we're part of the business school like. Like that's just that's the way organizations and higher ed institutions tend to co-mingle, right? And part of that is to your point, this supply demand thing, like this is what I think I'm gonna need from a skill set wise. This is what I think I'm gonna be able to produce. Let's bring those two things into parity. What we've been seeing, like we've been having a really fun conversation about the CPA which is a big exam that you take if you want to be a certified public accountant here in the United States. It has different variations in different countries and jurisdictions, but the upshot is there are certain requirements to being what's called CPA eligible, so you can sit for the board exam and all those good things. And we've been working with higher ed institutions as well as regulatory environments to bring those into what we think is a more accessible light and so and you can read about that or CEOs talked about that like there there's a lot there, but it's an example of like we've seen supply and demand mismatch, so we're trying to figure out how do we kind of nudge supply in the direction where we think demand is going to also be there. We've had similar conversations around data readiness and data intelligence. So, like, let's talk about how you get all tricks into classrooms and SPSS and SAS into classrooms so people can start doing big data analysis projects when they're juniors as opposed to when they're a 3rd year and they're dealing with something that's bigger than a pivot table, right? So that, so we've been doing this for a while. So, to some degree we have an established rapport, we have an established mechanism. What we're sort of still figuring out right now, and this isn't just us, this is everyone is OK, so like, wait a minute, I'm dealing with technology that's changing like quite literally by the minute. Like what if you type in ChatGPT something at the beginning of this call and type it in right now different answers because that's how fast the corpus is evolving, right? So, and to say nothing of the fact that the exponential nature of, you know, large language models. So, all of that is a mindset shift for both higher ed institutions and the organizations like us that help work with them to create the demand is we're now dealing with something that is both exceptionally variable and unplanned. So, we're not quite sure how this is going to play out in terms of our typical job architecture. Hey, I'm going to have a job as your first year. These are the tasks you do in an audit. These are the tasks you do on a on a functional transformation program. Like that's been fairly eddie steady, right, you know, steady for a while. Nope, like that's gonna change and it's gonna change faster than, you know, the economics department's ability to keep up with training people on those things. So, what we're trying to figure out is like what's the right balance between the specific technical skills I need built at the higher education level and the behavioral navigation skills I need built at the higher education level, that either you're building on purpose or your students are learning either on purpose or on accident.
And then conversely, what's our university talent acquisition team as they go into, into business schools and other universities, how do we suss out the individuals who are like really navigable, but really understand how to kind of navigate knowledge and understand how to kind of apply things that are new and show really strong learning agility, which is another skill set that's important to engender like. Our ability to suss that out is also being taxed right now.
Evan Metter: Let me, let me give a challenge to that which is, OK, so I've got my nephew's going to college. He's graduating here next month and he's going to go to college in the fall, and he specifically is going to a university that has like an intern, like a like a program like a work program, right? And when I thought you said, you know, about your kids and, and their education, liberal arts, I thought you were gonna say like college, like meh, like they're gonna have all this knowledge and etc. like colleges and inflation, college prices, it's gonna end up not being maybe worth it when you can acquire through MOOCs and other ways and that whole and we've had this debate, right? Like what was it that value of the knowledge, is it the brand, is it the network like where does the value that this come from? But my hypothesis back is like there's no replacement for experience and wisdom comes through doing. And so if I was thinking about it, this is my own bias, I recognize it is a lot more practical, um, education and internships and embedded types of work and I do have a hypothesis and I, I think this would be a loss in in our society, but I do have a hypothesis that the reaction is going to be much more practical applied learning like working that's gonna be trending away from a liberal arts, one, because it's such a culture war going on around universities, etc. and two, because of this economic fear of like relevancy and anything you can do to get a leg up from practical. So I, we'll see.
Louis: I see a lot of the conversation being centered around what are the skills that are gonna be needed to succeed with AI to actually embrace AI, how is that gonna shape kind of what we're looking for in new hires, but I think part of the conversation also is around, what are those skills or those abilities that might not be practiced as much when you're using and, I know we touched on this, but I think those skills actually, even if they're not as directly relevant with AI might still be relevant to practice somehow. And I'll relate this back to thinking something like we go to the gym, you know, a lot of people still work out, run, stay in shape. It's not necessarily something that we need to do to survive necessarily and, you know, we live in cities or we have cars, we have things that go a lot faster than a human running. But it's something that we need to do to stay healthy. And similarly, you know, when we had social media come up and, and kind of short form contents, a lot of people pivoted towards going back to reading books and actually listening to podcasts because it's more long form content and so it practices this kind of attention span that could have been lost with that more short form content. And so, I'm wondering if there's also skills that might not be as directly relevant in applying AI but might actually be really relevant to keep practicing somehow, almost like a muscle, you know, in order to be like a fully well functioning healthy member of society.
Evan Metter: Well can I, can I throw two ideas out there are somewhat challenging to me right now in this very moment. So one is your concept of empathy. I've been saying for like a long time, empathy is a 21st century skill and it's important, etc. So at the heart of empathy. It's about understanding and being able to feel someone else's experience, right? Habit someone else's experience. I am less sure now that GenAI is not going to be good at that. Like in sometimes I actually feel like some people's lack of imagination around tell me what someone else might be going through, right? Like I, so I, while I agree it's an essential skill, I'm not sure it's where we have a huge advantage over GenAI. I'm not, I'm not convinced of that #1. #2, I used to think that it's human trust, like people are going to trust, like it's our ability to develop a trusting relationship with people until Mike, I'm doing something internal where we're trying to change some things within how we operate in KPMG and you know, I'm writing white papers and I'm on conference calls and telling people like this is it and here's the vision and here's why it's going to be good. Do you know what really kind of humbled me was I just took a like a text of what I've written. I put it into like a GenAI thing that created a podcast, human people that weren't me, right? They basically told the organization what I wanted to tell them as third parties, AI that I scripted and it was more compelling. Like people found it, people trusted it more. Maybe I'm untrustworthy, right? Um, so I'm getting really, I'm searching. I wanna know like I'm not sure you can convince me that it is the human trust level, that it is our unique ability to empathize, but I wanna hear what your your reaction to that.
Mike DiClaudio: I think what you discovered was that the tools could accelerate certain aspects about what you want to do, in this case, get people on board with an idea. The idea nevertheless generates from you. And so the this we this human the loop phrase which is in like every AI publication you're gonna find is at the core what it's basically saying is AI can do a bunch of things, but it doesn't know which of them to do and in which order. So, you're the one that's really driving kind of when and how to insert it. So, if I were to play it back, I would say, look, what do I value? Do I value your ability to convince a bunch of people to do it differently, or do I value your ability to come up with the idea is? And then in an ongoing way, monitor, maintain, and compel. So, because if I value the second thing, then what you just told me was I have a really efficient way of getting after it. If I value the first thing, then yeah, I've just replaced you. But I'm not, I, so I think really the reckoning in the organization is, wait a minute, what is it I value again? Like do I value the effort, the process, or do I value the outcome?
Evan Metter: That's actually really that I feel is really profound. It's like vision and follow through, right? And there are things that that AI like those are two things that maybe AI doesn't. Doesn't do.
Mike DiClaudio: Like he was asking like what skills kind of atrophy that we need to hold like research is an issue, like just to be clear, because if I can type something like this is the, this is not the act of researching, like, oh, I knew the Dewey Decimal system and I learned that when I was in elementary school. Like I nobody knows what I just said, right? But my ability to research is now much more nuanced because I have access to the whole world at the moment, and the internet has always sort of granted us that, but now I have access to it, but I have to provide the context of the access. So I still need skills around what is a good way to research a problem, but the skill itself is different. It's different than knowing what publications to read or understanding and reading the long form to be able to drive an excerpt or like it's past that and it's much more into understanding media bias, understanding how to put all of these things into context, understanding how to play different research pieces off of one another to build a different point of view, like, so the research scale is going to change. I don't know if it's going to go away, but it's definitely going to change.
Evan Metter: Mike, I want to thank you for being our very first guest, like SNL with like the five timers Club. I feel like you're gonna be wearing that jacket, you know, sooner than sooner than anybody else but thanks very much and good luck out there.
