Episode Transcript
[00:00:00] Okay, Good morning, everybody. I just finished three weeks of travel in Europe, which was quite fun. Had a little bit of bike riding and a lot of work and going to various things. And while I was gone, a thousand things happened. And we're entering HR tech season and conferences. So what I want to do is kind of get you guys started on what happened last week primarily and then give you some long term perspectives. So two massive things happened last week. Number one, the job numbers came out, which I'll talk about in a minute. But even bigger than that, OpenAI pre launched some sort of a solution that they're developing for job matching, job placement and possibly hiring. And there's implications all over the place on what this means. So let me give you my perspectives on what happened and why they made that announcement. My guess is the reason they made that announcement is they are worried that about the impact of AI on jobs. And so what they want to do is get ahead of that issue and create an offering to improve the talent supply of AI skills, which by the way, is building now. And they don't really need to do this because there's plenty of other companies that do it, including us. But anyway, they want to get behind that and then they want to create some sort of a job matching system that allows AI talent to find AI jobs and AI employers to find AI talent. And they've got Accenture apparently involved in this, bcg, Walmart, the State of Texas and some other organizations. Now, there's a million ways they could do this. And let me talk about what those are in a minute. But first, why is this an issue? Well, this week we found out, despite Donald Trump's commentary about the BLS, that only 25,000 jobs were created in August. That's basically zero out of 165 million working people. That's less than 1/10 of 1%. So. So we've now reached a point in the United States where there's almost no job creation, perhaps zero. And that's not saying that everybody is looking for work and there are no jobs. Because if you're a front office worker or, or a retail worker or a hospitality worker or transportation worker, there's plenty of jobs. But in the white collar segment, which is about 30 to 35% of the workforce, it's shrinking. And there's two reasons why it's shrinking. One is AI. I know this because I've talked to at least 50 companies about it. Companies don't want to hire white collar workers right now because they've been told by vendors like OpenAI and Anthropic and others, that they won't need them anymore because they're going to be automated, including Microsoft, by the way. So employers, CEOs, CFOs are simply saying, okay, we believe you, we're going to stop hiring and we're going to spend our money on your tech. And so trillions of dollars, at least billions are going into these AI platforms. You've seen the numbers. Meta is building a, I think it's a trillion dollar data center the size of Manhattan, somewhere in South America. I believe we're investing in nuclear energy in the United States, data centers all over the place to accommodate this demand, quote, unquote, for AI coming from businesses and individuals. And so, you know, if you're a CEO or a CFO and you're seeing all this and reading all that and hearing all sorts of great stories about it, you're not going to hire a bunch of people. In fact, you're going to do the opposite. You're going to try to reduce the number of people you have, either quickly or slowly, or you're at a minimum going to freeze the headcount. And then you're going to say to your chro, figure out how to reduce the labor costs in the company and improve automation everywhere you can. And you're going to create a steering committee and you're going to start to do all sorts of projects. And that is exactly what's going on. And we have a whole model for this. I'm going to show it to you at the HR Tech conference next week and then the following couple of weeks in Europe and we'll tell you more about it. But that is what's going on everywhere. And in some cases, productivity improvements are happening in software engineering, in customer support, in hr. There are all sorts of things that AI does that eliminate routine work. So routine work, not routine jobs, Routine work. Routine tasks are being automated very quickly. And that gives you the opportunity to combine jobs and other things that I'll talk about in a minute. The second reason the job market is tanking is the tariffs. We don't know how much it's going to cost to import or export products around the world. The numbers keep changing, but we know it's not zero. And so more and more companies are simply waking up to the fact that they have to deal with this tax that the United States has created and therefore they have to reduce their margins and probably either reduce their growth path or their staff costs. So we do have a labor market slowdown now for job Seekers, as I've written about and did a whole podcast on, they're very concerned because we're all hoping to become AI gurus or superworkers as we call it, but we don't exactly know how. Even if we take courses online, they don't really teach you how to do your job, they just teach you about AI. So we're all going to have to get our hands dirty learning how to do this. And when we read stories about AI enabled tools replacing the need for a video editor or an instructional designer, or a sales analyst, or a financial analyst, or a lawyer, or whatever job it is you do, you get nervous and you look around and you worry about your skills. So the other thing that OpenAI discusses in this announcement is their hope to build online training for people that want to learn more about AI. So they're touching all of these markets that I spend my life in and most of you do, when they don't really know that much about it. They're an AI technology company, they don't know that much about recruiting, they don't know that much about sourcing, about skills matching, about skills assessment, about employee development, learning and development, career management, leadership development, all that stuff. That's not what they do. So these are wonderful announcements they're making. They sound really good. They sound exactly like what LinkedIn used to talk like 10 years ago when they were getting into this about helping economic opportunity and doing great things for society and reskilling people and all that. Well, look at what happened to LinkedIn. LinkedIn basically became an HR software company. Good for them. They're a big company, they're doing really.
[00:06:31] But it was actually a big shift. And you could argue that, you know, LinkedIn's core technology hasn't really been their focus for a long time. It's been all about building a better learning system, a better recruiting system and better tools for all that. So, you know, I don't, I would, I read the OpenAI thing as a very exciting announcement that we don't exactly know what it is yet, but what they're doing is responding to this crisis in the job market, in the market for job seekers, in the market for our careers and the economic issues. And I'm sure, sure Donald Trump is going to spin this in some positive way, but it's not positive for the economy because there's a lot of disruption. And while companies profits are doing fine at the moment, these internal transformation projects are very complex and slow and they're resulting in a lot of layoffs and disruption to Individuals. So I think the employee satisfaction data, the employee engagement data, the employee fear data and wellbeing data is plummeting. And I've written about this and I've seen the data. So that's what's going on now for you as an HR executive or an HR leader or a consultant. You're in the middle of this and we are all being forced to step up in a huge way. First, of course, we have to learn about AI and how to use it. And the number one recommendation I have for you is to get Galileo. We have built an entire system just for HR people and a version for managers. But let's just talk about the HR one to teach you about AI, to empower you, to give you prompts, to give you education, to give you training. We've connected our entire academy up to Galileo. So when you buy Galileo and Galileo Learn Together, you get 400 courses on all aspects of HR and management and technology. And you can access them as courses or through the agent by asking questions, you can create a tutor for yourself. We have built, I believe, the world's premier self development platform for at least HR professionals today. And I really, really encourage you to use it. But anyway, so there's that. Do we understand AI? Do we as individuals, do we as managers? Do we as leaders? Then does our company understand it? Do we have the right tools in place? Do we have an agent strategy that's rolled out so we don't have 15 different agents and hundreds of different platforms proliferating? Do we have a governance process and a steering committee to prioritize investments in AI? And then the big hairy part, this is the jobs. So let's talk about that now. The reason, and I'm going to show you some stuff on this at the next couple of conferences, the reason we have jobs is because they're collections of tasks.
[00:09:16] You know, if you look at job titles and job descriptions and there's hundreds and hundreds of thousands of them, the way they are created is we have a business process. Let's suppose we're selling some product. Maybe we're making clothes. And we have a bunch of steps we go through to make the clothes. We have to design them, we have to source the material, we have to cut it, we have to set up the manufacturing, we have to stitch it together, we have to finish the stitching, we have to dye it, we have to put it into packages, et cetera, et cetera, et cetera. And those are business processes. And we took the collections of business processes and we did organization design and we said, look, if we had a whole bunch of people specialized in this, we could do it more efficiently than if somebody tries to do the whole thing. So we segment these business processes into collections of tasks and we called it a job. And we didn't think that much more about it. We just said, look, you know, there's a job here to be stitcher, the stitchery person, or the dying person or the designer person or whatever. And we created more and more and more jobs. And when you look at the job change log database from Lightcast, every single day there are new job titles being created. So this is not a simple thing. When you read these, you know, kind of simplistic analyses of the job market, it's not simple at all. Every company has unique jobs created because of their business processes and their history in those business processes. Now, of course, because of the recruiting market and the job seeker market and all that, we've standardized many, many, many of these jobs. But along comes AI and it literally wipes out 30, 40, 50, 60% of the tactical work in white collar jobs. It doesn't change blue collar jobs that much. That, that's, but it will continue to happen. And so these jobs, which are collections of tasks, are getting regularly refactored by the AI. And by the way, the problem we have with it is not a one time effect, it's a continuous effect. Because the AI gets smarter by the minute. I mean, you can talk to AI, it can talk to you, it can analyze data, it can build charts, it can build graphics. I mean, it's going to be very human, like more and more. So, you know, when you define the job at scale, you have to define it in a way that it doesn't become obsolete in another six months. Well, what we've learned from the research we've done in org design for the last, I don't know, five years, is that the way to do that is to not focus on the tasks particularly you do have to do it, but to focus on the accountability of each role.
[00:11:56] What does this human being that we've defined this role to do? What is their accountability? What are they responsible for at an outcome level? And my best example I can think of is what's going on here. We have, we do a lot of content development, we develop courses, we develop videos, we develop research reports, graphics, PowerPoints, all sorts of things. And we have, you know, like most companies, a publishing group and they do all sorts of fancy stuff. And then we outsource all sorts of things to different people to build videos and things well, now that we can do this with AI, those tactical jobs have changed entirely to become super worker jobs. We just started building avatars this week. You're probably going to see an avatar of me in the next month. We've just started using digital twins to access people in our company who are not at their desk if you want to ask them questions. We're using AI for video generation, for course development in Galileo Learn. And so we can't predict how many things the AI is going to automate, but we can predict that it will continue to do more. So we're giving people broad job descriptions to learn about this and build their capabilities to produce more and more of these outputs that we want to do for you guys. In other words, our business processes without micro designing the jobs. So what's happening in jobs is number one, every job requires AI prompting. There's a big article I wrote this week in our website explains this. So all basically jobs have to include some skills in prompting, prompt engineering and understanding how AI works. But they're becoming broader and more contextual and more systemic and more business oriented. Because if you're maybe a pre sales rep and you know, in the sales organization and you're calling people to qualify them before a regular salesperson calls them, that used to be a very important job because we had no way of doing that electronically. Now we can pre qualify people electrically electronically, or we can pre qualify job candidates electronically. So the screener job more or less disappeared. So the person that was a screener, maybe their new responsibility is providing 10, 15, 20 pre screened candidates per day, per week, per month, whatever the number is, to the either hiring person or salesperson and they figure out how to do that using the AI. That's a very high level job, that's more of a managerial job. So part of this job creation problem is really job transformation. It's really reinvention. And that's what's going on inside companies. And we've been spending the last three months with about 40 companies in partnership with Rejig to talk to them about how they're doing this. And it's very confusing because everybody just built skills models or tried to for all their jobs and now they're saying, well, wait a minute, actually what I want to do is build a task model for all my jobs so then I can figure out what tasks are automated and where we should best apply AI. So you're going to hear more from me on this. But there's a flurry of new HR technologies being developed to show you tasks by job, to organize and centralize and analyze tasks by job to show you what tools tend to be good for what tasks, so that we can move beyond the old traditional collection of task, job creation, job definition, job description process. Now, most of you don't do this very often. And you know, a lot of companies don't even have a job architecture team, so they tend to do this fairly willy nilly. And what happens, that's a funny expression. I don't know why I use it, I've used it for years. But what happens is you end up with a bunch of jobs that mean different things to different people. And so, you know, if you look at your job architecture, pull up, pull out, out of, pull out of your hcm, a list of all the job titles in your company and all the job levels and all the pay levels, and you're going to see it's probably a mess. It's probably all sorts of duplicative things. Two different people doing the same thing with different titles, two different titles where they're actually doing the same thing. And that is the symptom of the fact that we didn't really have this job destruction technology around. So we would create jobs based on what was in the market and what people wanted to do. By the way, the other factor in job definition and job design is the candidate. I don't want to take a job that sounds stupid. It looks bad on my resume. So if you don't make the job look like a step up for me, I don't want to become the AI text analysis engineer or whatever you can, you know, contrive. I want to be the AI architect, you know what I mean? So this is very, very tricky. And those of you that work in talent intelligence and work in skills intelligence and in job architecture areas or talent acquisition, there's a lot of best practices here. And let me simply say the most important best practice is to keep it simple. Because if the job titles are simple and they're based on accountability, then you can morph what people do and they spend more time thinking about the job and less time thinking about how to get promoted to the next job. So fewer levels, fewer complex titles, broader definitions of accountability and responsibilities in jobs, and then you can build the training and learning and skills development that goes with it. And obviously the hiring and assessment. By the way, all of this complexity is going to land in the lap of those guys at OpenAI. I don't think they've really even thought about it yet. Maybe they have. And then the Last thing that comes out of this flurry of news about OpenAI and the job market and the money being spent by different vendors on AI tools and technologies and the shift in economics is learning.
[00:17:52] Now, most of you know I've spent the bulk of my HR career in learning and development. I've written three books in the learning development area alone. And we have to rethink that too, because the way learning and development was created, it was patterned after jobs. We have job definition, job title. Let's build a training program for that job. Well, you know, that never worked super well because each person coming into that job comes with a different level of experience already and they may or may not need the details, or maybe they need more details on how to do the job. There's always the issues of how the job actually gets implemented, not how it was theoretically designed. The changing nature of the job from tools and business processes that change at the front line in the work experience itself, the role of the manager in coaching and developing people, and then the soft skills, the human skills, the complex thinking skills, complex problem solving skills on the job. So, you know, it's always been very, very hard to build training around one job. So we tended to train people on topics, on systems, on programs, on skills, because the job itself was always kind of too amorphous and too complex. Well, now that we have AI and we have Galileo Learn, let's flip that and focus it on the individual, not the job. Let's change the whole paradigm and imagine that we have a learning system that has the corpus of knowledge of how we do things in our company. That's basically what AI does. That's what we did with Galileo. And let's suppose as a learning and development HR team, we are regularly interviewing subject matter experts, getting information from product people and line operations people on business processes, coming up with standards for procedures, making sure that regulatory and compliance data is clearly articulated, that our policies are up to date, and we're continually continuously talking to leaders about their business strategies, competitive strategies, what to do about changes in the economy, what countries we're going to do business in, what's important to us, what's not important, and so forth. Let's suppose we have all that information, which we do, it's sitting around in millions of different places inside of companies. And we stuck it into this AI, Galileo Learn platform, or whatever one you buy, but there aren't that many yet. They're coming. And you then said to the employee, you have a self development tool or system or platform that we will deliver to you in the flow of work to learn and develop and advance in the role that you've been given today. Now you might say to yourself, that'll never work because people don't have time to take training and they don't have time to ask a bunch of questions and we need to give them more formal education.
[00:20:42] And I would argue that you're wrong. Look at the growth of ChatGPT. ChatGPT went from zero to a billion users in one year. No learning program has ever come even within, I don't know, maybe 5% or maybe 1% of that growth rate. Because we learn through curiosity, we learn by asking questions. We learn through mentoring, we learn through coaching. Sometimes we want to read something, sometimes we want to watch a video, sometimes we want to look at a diagram. Sometimes we just to talk to somebody about something, sometimes we want to challenge something. You can't do that in a training training. You could in physical training, but we don't do as much of that now. So in addition to the online stuff and the virtual training and everything else we do, we're going to all have learning agents that teach us stuff. And that's what Galileo is. Galileo is not just a problem solving agent, an automation agent, a data analytics agent, a benchmarking agent. It is a learning agent. It's going to be. It already is, because we've already built this tool to help you develop.
[00:21:45] I'll tell you a funny story that happened to me. I come home from Europe late yesterday, wake up at 3 o' clock in the morning today because I'm on European time zone. My Internet's not working. I hate it when that happens because I got to go into the wiring closet and figure out what's going on. So I go in there and I look at the Xfinity router and this and that and the other thing and I find out that one of my Apple routers has died. And Apple doesn't sell routers anymore. So it's dead and I'm not going to replace it. I got to figure out how to work around it. So I take a picture of the Xfinity router and I stick it in Google and I ask Google, what the heck is this? And it gives me a model number. I go to ChatGPT, I tell my little friend over there what's going on, that my computers aren't working, but my WI fi is. And I give it the router number and it starts, it says, oh, I know exactly what that router does. It has this feature, this feature this feature, this feature, check this, check that. So I go back to the router, I try to log into it. I can't log into it because I don't even know how to log into it. So I go to my Xfinity customer service portal and I get a little agent that comes up and says, type this DNS into your browser and log in. And I type the DNS and it says, admin password, login password. I don't know the login password. So it teaches me how to do that. I log in, I look at the back end of the router, I go back to ChatGPT, I say, I found it, here's what's going on, blah, blah, blah. And literally within 15 minutes, I figured out what to do. Now, do you call that learning? I don't want to take a course on Xfinity routers. I don't want to take a course on the Internet. I don't want to take a course on the Apple machine versus the other one, versus, you know, the Xfinity one. I mean, I really, really just want to solve this problem and I am going to get smarter as a result. And by the way, while I was going through that exercise, I was asking ChatGPT a lot of questions about what is a NAT NAT and what is IP conflict and so forth. It was teaching me this stuff. And then I'm done. I got my problem solved. Why wouldn't we all want to live like that? Now, there are times when I'm going to say to myself, you know, I really would like to understand this. I'm willing to spend a half an hour, an hour with somebody, just explain to me this whole thing so I don't have to go through it again when I upgrade to the next router. But that's about enough. I don't need a curriculum on the Internet and all of the implications of different DNS, you know, strategies and this, that and the other thing. I don't need that right now. Maybe it's not my job to learn that much. So you can see that the self development learning agent idea is massive. Massive. Now, OpenAI hinted about it, they're working on it, but we've done it. And I'm going to make a little bit of a plug that if you want to see what this looks like, just reach out to us and we'll give you a demo. You can do this today. And so of all the AI transformations you're dealing with in your company, and there are hundreds, maybe the biggest is how we teach people and enable employees to learn. Because if we can't do that, how are they going to keep up with all these other changes being created by AI? And so we now have a good number of clients doing this with us. It is a revolutionary way to think about learning and enablement and support for your employees. And it plugs into your existing AI infrastructure just fine. You don't have to start all over and get in an argument or a debate with your IT department about what you want to do. So that is the sort of third big thing that I want to talk about this week. Okay. A little bit too much personal stuff, but I thought you guys might get a kick out of that. This week, I'm in Vegas. Next week for the HR Tech Show. We'll be at SAP Connect launching a big project we've been working on for more than a year with SAP. We will be at Unleash in Vegas and also at Workday Rising in the US and probably in Barcelona to show you what we've been doing with Workday. And I'm here to answer your questions, talk to you about these projects, teach you about the AI journeys that we're going through, and we will be teaching you more about the super worker effect in companies everywhere we go. This is all very complex and confusing and rapidly changing, but it is one of the most fascinating opportunities you have in your career right now. So I encourage you to just do what I did today with my router and just learn, learn, learn, and you're going to get smarter and smarter and smarter about all of this, and you're going to be very successful as a result. Thank you.
