Episode Transcript
[00:00:04] Hello, everyone, it's Josh Burson. I am recording in Stockholm today, where we're visiting Sana Labs. Just finished three days at the unleashed conference in Paris, and I want to spend this podcast talking about technology and AI and what's going on in HR at this particular point in time. So I talked to a lot of clients, clients up here and a lot of vendors, and got a lot of feedback from my keynote on lots and lots of topics. And the overriding message I like to give in this podcast is this AI stuff is real. It is working, and it has massive, massive impact on what we do in HR, to say nothing of what it's going to do to every other job in every company. And if you want to understand it, you have to get your hands dirty with it. So we had a lot of meetings with companies asking various conceptual questions about what's going to happen to recruiters, what could happen, what's going to happen to business partners, what's going to happen to call centers, what's going to happen to L and D, etcetera.
[00:01:10] I can probably tell you answers to those questions, but the way you're going to find that out is by using these tools and understanding what parts of your job as an individual and your workflows as a company can be reimagined. And I mean reimagined because the power of these AI tools is so beyond your expectations and sometimes imperfect that you won't really know how to apply it until you use it yourself. And I don't mean yourself as a company. I mean yourself as an individual. And by the way, this is the one of the biggest use cases for Galileo. Galileo is a full function, very high powered AI platform, but also an information assistant. And people that use Galileo learn what's possible with AI in training, in change management and knowledge management, in design, in recruiting, in L and D, in customer support, because Galileo can do all of those things. But you have to see how it works to understand it. And I'm not going to give you a tutorial on what it is at this point in time, but this stuff is so powerful, I can't explain it to you in a sentence because it does so many things that will only be unique and useful to you based on what your issues are and where you want to focus your attention. Now, I went through probably 50 use cases for the audience that unleash, and we're happy to go through this for you guys. We now have four courses on AI in the Josh person Academy, and more and more content being added to the Galileo success Center. So you can get examples and real specific prompts and tools to figure out ways to use AI in many, many areas of HR. And we're adding more content to Galileo, too. So we're going to go into compensation, we're going to go into location, strategy, and a lot of other things that will help you figure out where AI fits. The big kind of new issues with AI are really three big things. First of all, the large platform vendors SAP success factors, in particular workday. And at some point, Oracle are adding underlying AI data platforms under their erps. And what that means is that the data that you create or use or analyze in one isolated part of your company, like recruiting data, learning data, career data, salary, whatever it may be, can now be stitched together and correlated and analyzed in the context of other business data. And in some sense, this may be the first time in 30 years that we have a revitalization of the concept of ERP. The concept of ERP, which is a terrible name because it stands for enterprise resource planning, but that's not really what it is, is that with one platform, not only do you reduce the number of vendors, but you can ask questions about financial issues and understand supply chain, marketing, sales, customer HR, staffing, personnel issues that relate to that financial issue. So if your company is behind on its quarter, and maybe its profit, profitability is behind plan or revenue, whatever it may be, the system using AI can dig around in the ERP and find the anomalies that are contributing to that problem. That actually sounds so simple, but most of these systems were never really very good at doing that. You had to buy an analytics tool on top of it to even figure out how to do that analysis. Well, AI can do that for you. Now, within HR, of course, we have the same situation, turnover related to pay, related to DEI, related to tenure, related to manager, related to location, related to a hundred other things. How do we find a leader in the company who can handle this new project when we don't have a good hipo process and we have too much bias? Who needs new skills? How do we best develop the new skills? How do we keep the skills, data and content up to date? I mean, I can go through this for dozens and dozens of use cases. This is a completely different world we're entering in AI now. It's not going to all work day one. Those tools are kind of getting developed in real time, but this is a massive, massive rethink, reimagine how everything works, and I kind of mean everything. Call centers, business partners, recruiting pay systems, the way we do diversity, the way we do learning, the way we do career management, the way we do internal mobility, the way we do job scheduling, the way we do hourly scheduling, all of those things that we do in HR, including employee relations, are going to be aided and improved with various AI solutions. And actually, I'll bet a year from now we don't even use the word AI anymore. They're just going to be new tools and that'll be it. We don't need to worry about how they work. But today, this is what's happening. And the best way for you to understand how to do this is to get your hands on these tools, play with them, use them as a team, experiment, do some hackathons, and start to implement some changes. And it's very clear to me, with the agentic AI, the agents that are being developed now by vendors, you're going to redesign the way you do a lot of things in HR, simplify teams, reduce or eliminate a lot of routine work, and probably shrink the headcount in some areas to create more opportunities for HR people to do what you want to do, which is advise and counsel and support managers and leaders in the company, and not get bogged down with internal HR stuff. Now, in addition to that, there is this new emergence of active agents and tools. Many, many vendors are demonstrating what they might call agents or copilots. And the idea here is that in addition to the large language model, which can reformulate and generate language or content or graphics or images or videos, we can also have a large action model. So the LLM, I guess, is what we call it, has action frameworks within it and can take action based on prompts or input and from you. So if you think about workday, for example, or SAP, I'll just mention those two. There are thousands and thousands of transactions possible in those systems. They are hidden and buried and nested into panels and screens and workflows so that you don't just get to the transaction by itself. You go through some process to get to it, to make sure that you're doing it correctly and the right data is entered into the system. The AI can theoretically access all of those transactions through APIs and bring context to them. So instead of going through the 15 screens it takes to enter your vacation request to your manager, or whatever it may be, you simply say to jewel or the workday assistant. Jewel being the SAP assistant, I would like to take a vacation in three weeks for four days. What is my balance? And would you please ask my manager for approval. In fact, if you're using the Microsoft copilot, you could probably ask it what is on my calendar that needs to be rescheduled. And the system can do that in an integrated way if it is agentified or designed to be an agent. And all of these tools are doing that. We're going to be doing this with Galileo originally. Initially, we're starting with workday. With our integrations with workday, we'll do it with other systems.
[00:08:38] All of the user interfaces that are sitting on your portals or websites or whatever it is that you're using to support your employees are going to be very, very active. They're not just chatbots. Chatbot does not do it justice, really. It's much more powerful than that. So I think the word agent is probably the right word. And you're going to be able to design these things because there are tools from Microsoft, from Servicenow, from SAP, probably eventually from workday to allow you to create these agentic workflows yourself so you don't have to wait for the vendors to figure out which one's the most important. You're going to be doing this, and I don't think it's going to be doing it. I think HR is going to be doing it. I think these are going to be fairly easy to use interfaces. So that's number two. The number three thing in AI that came up a lot this week that I think is absolutely fascinating is what is called a digital twin. Now, when I first heard about a digital twin, somebody was telling me about it a year or two ago. I thought it was like a game where, you know, I would have be. I'd be first person shooter, and there'd be another first person shooter, and my digital twin would be, you know, shooting with me or at me or whatever, I was really wrong. What a digital twin originally was used in manufacturing was to simulate a manufacturing plant. So before you go out there and build a manufacturing plant, you can simulate it digitally and see how well it works to deal with issues like supply chain and speed and process manufacturing and so forth. Well, in the context of white collar work, a digital twin is an agent that is like you, that knows what you know in some sense. Galileo is a digital twin of me. It has my voice, but it also has all the research I've ever done, most of it, not all of it, but most of it, and most of the research that everybody else has done around me. So it's actually a little bit smarter than I am. But if I added to Galileo, all of my email traffic for the last ten years, everybody I've communicated with and everything I've said to them and what they've said to me, all the documents on my computer, all the meetings I've been to, and the interactions between them. Can you see what that corpus would be? That corpus essentially defines my personal work life for the last decade. And if we added to it a bunch of business process or business rules, we could turn this digital version of me, and I'm just using me as an example into a super salesman, a super HR leader, a super analyst, a super call center agent, whatever you may want to create. In fact, one of the insurance companies, one of our large clients, now has digital twins as claims agents. Because processing claims is very complex. It takes a lot of time. There's a lot of back and forth with the client getting information on, you know, receipts and spending and various parts of the claim, assessments of damage and things like that that can be digitized using expertise from humans and process work. So now we have an agent that can keep track of the claims process for thousands and thousands of customers. Now, I'm not saying the customer would interact with the agent directly. Maybe a human being interacts with the agent, or the agent supports the human being. But I think we're going to have a point very soon, and I'm talking about probably next year, where most of us are going to have a digital twin who's helping us with our work, and that digital twin will probably be able to help others with their work when you're not around. Let's suppose you go on vacation for two weeks, or you have family leave, or you take a year off. The digital twin, which, by the way, is probably owned by the company, not you. Most of your work product is owned by the company, can essentially help continue the business operations while you're not around. Now, I'm not saying it's going to learn and interact and maybe be as, you know, judgmental or experimental or empathetic or whatever it is that you're good at, you know, than you are. But it might be, it might have picked that up from you, too, because if it really looked at all those interactions, it would know a lot about your style, and it could mimic that style also. So in some sense, these are the optimist robots of white collar workers, and we're going to all have maybe multiples of these. Maybe we're going to have one at home to help with the shopping and the vendors that help you clean and fix and repair things around the house and you know, how do I reset the sprinkler that I keep forgetting how to use? And where did I put the such and such, and what drawer did I put so and so? I mean, all these sort of informational related things that we do at home probably belongs in our personal digital twin at home. And there'll be digital twins at work, and they'll be specialized digital agents that do other things. So, you know, this idea that has been floated for a long time that we're going to have digital employees is actually becoming true, as odd as it sounds. Now, I'm not suggesting you rush off and try to build digital twins right this minute because the vendors haven't quite given us solutions for this, and I know quite a few of them are coming. But I want you to understand the concept so that when you use systems like Galileo or the Microsoft Copilot or the agent from Servicenow or the workday assistant or Juul or whatever it may be, you raise your expectations on what the system is capable of doing and ask it to do more. And this is really one of my big learnings about Galileo. People aren't pushing it hard enough. They're not asking it enough. Rather than ask it a simple question like, do you have any research on recruiting? And the answer is obviously yes. Give it a very, very specific request or prompt or demand or inquiry, whatever you want to call it, and tell it what you want to do and ask Galileo or the other agency you're using to do it for you and see what happens. Now, it won't always work. Sometimes it'll make a mistake or it'll be missing some stuff. In our case, we're very carefully trained, Galileo, not to make things up. Sometimes it will make things up, but that's where this is going. And, you know, this was something that's been brewing. I've been thinking about it, and this week I talked to some customers actually doing it. We will be doing case studies on these. If you are building digital twins or using agents in your company in any particularly interesting way, let us know immediately. We're going to do a podcast on you. I'll invite you onto a webinar. We'll write a case study about you. I really want to get this information out there and try to explore the use cases of this because this is really, really exciting. Okay, a couple more quick things while I go through what happened this week.
[00:15:28] We are far along now in our research on AI and learning and development.
[00:15:34] Probably in Q one. We will have this all published and I show you what we've learned. If you are thinking about learning and development, reinvention, reimagination, please call us. We are way down that learning curve because we're doing it ourselves in the JBA with Galileo so we can tell you what's possible.
[00:15:54] And that particular topic is our core. We probably have more expertise in learning and development than anything else. So if that happens to be on your mind, just reach out to us because we really want to work with clients on this topic and hear what your issues are and help you put together a plan and do some implementation of what you'd like to accomplish in the area of recruiting. Had a bunch of interesting conversations with clients.
[00:16:19] Strange thing happening that now that we have high powered, creative, generative AI on the recruitment side and also on the candidate side, candidates are using AIH to read the carefully crafted job descriptions that you're using AI to create and the skills that are implicit in those job descriptions, and they are mimicking that to create perfect resumes so all the candidates look the same. Oh boy, that is a problem. So all of a sudden we have an arms race between the candidate and the recruiter, and the candidate has access to the same technology as a recruiter. So you think you're scoring from higher school or whatever it may be, or, you know, whatever tool you have is helping you screen out candidates. Maybe it's not going to work anymore. Maybe it's not going to work at all because everybody's going to look the same.
[00:17:11] So what the client I was talking about with this about said, and it was a really interesting idea, we actually talked about it for an hour, is maybe we need to make it harder to apply for a job, not easier. We keep the last, I don't know, decade we've been trying to make it easier. Maybe we should make it harder. Maybe we should ask more questions. Maybe you should make the candidate interact in a human way. Maybe there should be a few captchas here and there before somebody submits their resume.
[00:17:39] Maybe we should do more screening and assessment before we allow them to create a username or put their resume into the system. And so that's a really interesting topic to think about, because even though the job market is very competitive and we have low unemployment and a lack of skills in a whole bunch of areas, you want the right person. And if you're going to be flooded with candidates that don't look real or all look the same, it's going to make it even harder to find the right person. So I'll keep you up to date on what I've learned in that area and again in that area. Just let us know if you're just doing something interesting we'd like to share with others. The final topic I want to mention briefly is energy, electric energy.
[00:18:21] I think a lot of you probably know my academic degree is in mechanical engineering. I have a bachelor's and master's in mechanical engineering. I studied energy in the seventies and the eighties. That was what I was going to do with my career back in the oil crisis days.
[00:18:37] And all of a sudden it turns out that all of this wonderful AI that we use to reschedule meetings and transcribe things and all the things I mentioned earlier takes a lot of energy. And it was really kind of laughable to me that this week, in addition to Microsoft signing up to buy most of the power from three Mile Island, Amazon invested in two or three nuclear, mini nuclear power plants to build electricity for their systems. So the funny thing about all of this AI stuff is it is colliding with global warming. So on the one hand, we have this sort of circular social system going on wherever people are having fewer children, we have delays in the marriage rate. So the workforce is getting smaller in the developed countries, the unemployment rate is going down. So we're going to have shortages of workers. We have AI to help us improve productivity, reduce the number of workers we need, presumably, and help us improve productivity and growth in a world of fewer workers. And the AI generates heat from electricity to high demands of electricity, which in turn creates global warming, which in turn makes our lives worse because we have migrants moving all over the world and hurricanes and fires in California and a whole bunch of other things going on which makes people depressed so they have fewer children. You see, this is kind of like a big circle.
[00:20:09] I mean, maybe this is a silly analogy, but, I mean, I suddenly saw that all these weird things going on are all connected together. And then right in the center of it is this horrible political process we have in the United States and other countries in the middle of it. I don't know what to do about all that, but I just wanted to point that out because I thought it was kind of interesting because we're a part of this. We're part of the labor market, we're part of the AI revolution. We're part of the productivity issues in companies. We're part of the culture of the workforce, were part of, you know, really helping people understand why they're unhappy or disgruntled with their jobs and their lives and other parts of their family's. Issues, providing them benefits. And it's really interesting to me how this is all sort of tied together in this gordian knot of the coming decade. So, anyway, I have a lot of other thoughts to share with you, which are going to come out later this year when I prepare the predictions. That's kind of an update from Europe on AI. It is all good stuff.
[00:21:12] These are fantastic tools, really, reinvention tools. We're going to have quite a bit of reimagining in 2025, and I will tell you more about that in the next couple of months. Get hold of Galileo. It's only $49 a month. No dollar, 39 a month. And you can learn everything you need to know about AI right there. Get your hands on it, and I promise you, you're going to be really happy that you did. Okay, have a great weekend, everyone. Bye for now.
