Episode Transcript
[00:00:00] Speaker A: Good morning, everybody. Today I'm going to go through three topics, very contemporary and kind of urgent topics, relative to where we are in the world of HR and employment and AI. The first is I'm going to talk about the vendor market around AI. Then I'm going to talk about the fleeting attempt to find returns on investment and why that is happening. And then I want to do a slight plug for the irresistible conference in
[00:00:27] Speaker B: June and tell you what we're going to be doing there. So.
[00:00:29] Speaker A: So, first of all, the spectacular thing that's happened in the vendor market of AI is it's everywhere. We're all using these tools. They're embedded into our phones, they're embedded into our computers, they're embedded into our apps. There are agents being developed in every part of hr. Every HR technology vendor has an agent or is building an agent or a series of agents. And many of the things that were considered to be quite sophisticated and quite difficult to do are becoming comm.
So what we're going to be doing to keep you ahead of this is writing a series of articles on the different parts of HR and how many vendors are changing their strategies and their tools for you to build internal and embedded agents. Now, the problem with this for you guys is there's too many things out there and you've got to be very selective. And I'm going to talk a little bit about that. And our blueprint is designed to help you think this through. And when you listen to the part of the podcast about roi, you you'll hear a bit of a story as to why and how you can make sense of this noise. But the commoditization is rather incredible. For example, something that was so unique and weird and interesting a year ago, the ability to take a document and turn it into a podcast, to take a document and turn it into a course, to take a document and turn it into a video, to take a document and turn it into an agent. You can buy that from a hundred different vendors for maybe $10 a month, if that.
So the generation capabilities of taking any form of content you already have, policies, courses, videos, audios, speeches, you know, regulatory rules, whatever, and turning them into an AI object that can be queried and chatted with and converted to training is. Is essentially everywhere. It's free, it's not quite free, but it's pretty close to being free. So if that's why you're buying a tool, be careful, because that isn't enough. What's going to matter at the application layer? And application layer is where all the value is in AI, by the way. The infrastructure and the compute and the electricity and the data centers are probably the most commoditization of all. Even though there's a massive amount of money going there, is how the application works, how well trained it is, how well it understands your industry, your company. And a lot of that falls on your shoulders, not just the vendor, but you'll see that the vendors have many, many different tweaks and tunes and angles to this. And so the first time you see one of these generation tools, you might get astounded. And then you're going to realize, well, maybe we need to find the one that's the best for our use case. For example, in the dynamic enablement market, which we're spending a lot of time in, are we enabling frontline workers? Are we enabling engineers? Are we enabling support people?
Re enabling executives, salespeople? Different use cases require different implementations of the technology of dynamic content generation. Dynamic content generation itself is not really the feature anymore. It's how well does it serve the needs of the audience that we're trying to address. And that means that the vendors have to be focused on problems, not tech. So, anyway, that's number one. And the podcast chapter I'm going to insert into this talks about how you think that through. Second thing is a little bit of a plug for Irresistible and what we're doing with Galileo. Now in the next two weeks, there's going to be a series of HR conferences here in the U.S. unleash and transform, and we're going to announce the new version of Galileo and a relationship with another vendor that really moves Galileo into the world of completely digital hr business partner available to every employee in your company. And so we're implementing essentially a strategy we've been working on for a couple of years, which I guess I would call Galileo everywhere. Galileo as an agent. Galileo in workday, Galileo in ServiceNow, Galileo in Copilot, Galileo in joule, Galileo and all in Claude, in ChatGPT, in whatever AI you have, Galileo in Vendor X, Y, Z, Hibob, et cetera. The purpose of that is to allow you to access the intelligence that we've gathered for the last 30 or 40 years in whatever systems you have. Because AI is everywhere, and it is a almost standard technology now in enterprise tech. So we have found ways with MCP and other technologies to integrate Galileo into all these other systems.
[00:05:10] Speaker B: So that's coming up.
[00:05:11] Speaker A: And then on June 8th through 10th at USC, see, we're going to have our 5th annual Irresistible for those of you that are leaders or global HR professionals or CHROs, I urge you to come. We're going to be introducing a new leadership offering for you. We're also going to be showing you how all these things work. We have some amazing speakers coming. We're going to talk about AI governance, AI deployment and change management, and technology and strategy from some of the world's largest airlines, including iata, the industry association of airlines, some large pharmaceutical companies, some large retail and consumer goods companies. So you can really learn a lot in those two and a half days. And if you bring a team with
[00:05:54] Speaker B: you, you will be able to have
[00:05:56] Speaker A: an analyst from our company assist you as a team and give you a wrap up of the conference so you can go to different sessions and come back and consolidate what you've learned and then bring it home as a opportunity to refine your strategy. Okay, so that's a little start here. Let me now cut over to the discussion of the fleeting ROI of AI and how you really put a business case and strong business strategy behind all of this technology. Bye for now.
[00:06:28] Speaker B: Today I want to explain the challenging situation going on in the world of AI, where billions of dollars are being spent on tools, billions of dollars are being spent on data centers, electricity production, real estate, fancy semiconductors, and the return on investment in the business cases. And the productivity numbers are not showing up. Now, I'm not going to quote all of the studies that have been done, but there have been maybe 20 studies that I've seen that have shown that the actual impact of AI on the economy and on companies is pretty small compared to the massive, massive investment. In fact, in terms of the size of the investment, a lot of the economists believe that the ENT growth in US GDP has been 100% driven by AI investment. And the job market, which has remained relatively stable, is being buoyed up by jobs in frontline work and healthcare and other things, while companies try to lay off white collar work but haven't really been effectively doing it well. There's really two reasons this is happening and they're not hard to understand.
Number one, it's not clear to most leaders, and I know this because I talk to a lot of executives where to apply AI. There are hundreds and hundreds and hundreds of opportunities to apply AI in a business. You can give it to the accountants, you can give it to the salespeople, you can give it to the marketing people, you can give it to the engineers, to the software people, to the IT people, you can try to replace your mainframes, you can try to replace your customer systems, you can try to replace your sales systems, you can try to replace your websites. It's almost like there are too many moving targets. Probably a little bit like what it feels like when you look over Iran when their military bases are gone. So of course what happens is people try a lot of stuff. So there are many, many, many experiments going on around companies with AI software, engineers, marketing people, salespeople in the supplier market. Then a market is doing a great job of producing half built solutions for all these things. You can go to a marketing tool, a training tool, a recruiting tool, or you know, a content development tool and you will find an AI feature already there that looks like it's going to be 10 times better. But that feature is not really that well built out. It hasn't been refined to the point that it's clear exactly how it's going to change your company because the vendor doesn't know either. So we're buying a lot of this stuff. In fact, I think the money being spent on AI this year is going to total up to be a very large number at the end of the year. But the tools are not well refined. So we have lots and lots of small experiments. And what happens in a large company or a small company is lots of small experiments don't result in a big return on investment. Because as many of you know, we've basically built companies with lots and lots and lots of complex job titles, workflows, legacy systems and organization structures that are hard to change. And if you don't have a clear picture as to how the technology will work, sprinkling or spreading peanut butter on all of these jobs and hoping that they're all going to get better doesn't really result in a huge return. Now we know that in recruiting, for example, there are at least 12 to 15 steps. And there are sourcers, there are schedulers of interviews, there are people that place ads, there are people that post job descriptions, there are people that study the job architecture, there are people that create interviews and do interviews. There are people that manage the talent acquisition technology, there are people that do analytics, there are people that look at onboarding. I mean, there's lots and lots of people that do a lot of things just to bring large numbers of people into a company. If you don't have a streamlined system like Paradox, like Maki, people like Efold that does a whole bunch of those things, you will automate a piece of them, but the interconnectivity between the pieces won't be very high. And so Executives look at these opportunities and they say, let's apply AI to recruiting. But then the details fall down into the operations teams and they look around and they say, well, what does that mean? Do I buy 20 new AI things? Do I buy one? Do I build it? Do I wait for my core vendor to automate it? Do I turn on all the agents that they have inside the core system? And by the way, this is another problem is that the big HR software companies, Workday, Oracle, SAP and all the others are adding AI features to their current systems and selling the productivity improvements of each feature. And those are, you know, not zero return use cases, but they don't change the process that much, so you'll see some return. But if the recruiters are doing the same work, maybe they do it 30% faster, but the process of recruiting may not be that much better and you may not see that much return. So we have these, you know, sort of all sorts of small incremental changes taking place with the downside that it takes time for people to learn how to use the tools. You don't know exactly the impact is going to be, and the company can't measure the return. I've had three or four major meetings this week with companies who said to me, we've bought tool A, tool B, tool C. Hear from people that they're doing more work, getting more done, but it's also making their jobs busier.
[00:12:14] Speaker A: Because if you get your hands on
[00:12:15] Speaker B: a fantastic tool, you can iterate faster, you can do more things, but will the result actually be more efficient just because you're doing something faster? Maybe the thing you're trying to do shouldn't happen at all. So executives have a really hard time deciding where to focus. And the reason for that is they don't necessarily know how mature these different use cases are and where to apply it. Now we know that in hr, the five big return areas are very clear. This is all part of our blueprint. Number one is employee services, employee self service. Massive, massive return here for employees, not just for the call centers to make employees more productive. Number two is in talent acquisition, where you can reduce many, many steps and improve hiring speed, quality and strategy, not just reduce the number of recruiters. Number three in L and D, where you can move from static training to dynamic enablement, which is a massive roi. Number four is in the HR business partner or the embedded HR role, where you can literally replace many, many of those steps and tasks and roles with a digital HR business partner like Galileo, which we're going to be Spending much more time talking about and move those people up a level into more strategic advisors. And the fifth is in the automation of analytics, data management and pay equity and other data related things that HR does where we don't need a whole bunch of analysts figuring stuff out by hand because the agents can do that too. All of those areas have very high return on investment. But executives have to decide where are we going to focus, which is going to be our top priority. Are we going to focus on onboarding, Are we going to focus on sourcing, Are we going to focus on high volume, Are we going to focus on executives, et cetera? Number two, the second reason this return on investment is so hard to measure is because the technology is so new. The things that AI does are spectacular, there's no question about that. I mean, we use it every day, but the actual functional activities keep changing. Within almost a week at a time, you find features in the AI that didn't exist the week before. And so what happens is whoever is implementing some tool or some system is frustrated or perhaps concerned that the thing that they're doing isn't keeping up. And the vendor market is extremely competitive. In fact, I think a lot of the things that we consider to be highly proprietary and very unique are quickly becoming commoditized. Let me give you two simple examples. Skills inference. Now, I'm going to publish a really interesting interview with Ashutosh Garg, the founder of Eightfold, in a week or so. But skills inference was a very arcane, difficult, complex technological problem.
Ten years ago we had, you know, early tools trying to do this in applicant tracking systems. Virtually nothing in the learning industry other than assessments. And then along comes Eightfold, who basically invented a bunch of AI modules in their system that can infer skills from job descriptions, from job history, from career paths, from certifications and other things. And at the beginning of that space, people were spending hundreds of thousands of dollars or more per year on skills inference engines. Companies like Lightcast were born that built a massive data set and standardization of skills. Skyhive, which is acquired by Cornerstone, techwealth, a whole bunch of others, in fact there were dozens. Some of them have been acquired, some of them have gone out of business. And we believed, and most HR people believe, that this is going to be a huge technology workday, invested in the skills cloud and purchased a bunch of companies to try to get that to work. SAP launched a big skills and intelligence project. Oracle claimed to do the same thing and so forth. And most of you and most companies bought skills based Technology, by the way, a lot of this started from degreed, who tagged skills in the LXP manually, by the way. That wasn't really AI driven, but you know, Ed Cast did it and others. And we suddenly said, great, let's go spend a bunch of money on this.
[00:16:37] Speaker A: Because if we know the skills of
[00:16:39] Speaker B: all our employees, we can better hire, we can better develop people, we can better move people internally. We had the development of the talent marketplace market, et cetera, gloat, all of that. Fast forward to today. This is a commodity. You can take a job description or a recording of a meeting or a series of emails and stick them into ChatGPT or Claude or Gemini and it will tell you pretty accurately what the skills of this person are. Now what that means is that if you went down that path and barked up that tree, you're looking at technology that's almost obsolete already. Because now skills technology is in everything. I mean, it is, it is in everything. Galileo is a very, very, very sophisticated skills inference engine in and of itself. And I've talked about this many times. So you know, if you tried to do that five, six, seven years ago and you spent a bunch of money on it, in particular a big project, her CFO and Chro is scratching their head and saying, well, you guys already got a bunch of money for this, why do you need more? We already spent X, Y, Z on this. Where's the return on investment?
That has been a frustrating experience for many, many companies. I, you know, I, I advised a lot of people not to do this, but it happened anyway because the market was so exciting. And there's other examples of that. AI based interviewing, which started a long time ago with companies like Hirevue, is now becoming a commodity. AI based coaching is becoming a commodity. AI based content development for training is becoming a commodity. So these rapid changes in technology are making it difficult to generate meaningful projects with a return on investment when you spend a bunch of money on something else. So how do we change all this? You build a blueprint, you build a roadmap, you strategize using help from companies like us to figure out where the near term high return on investment opportunities are for your company. Not for everybody, but for you.
And that process of refining the applications for your technology and deciding where to invest based on the maturity of the technology and the opportunities in your company is the same process we have used for mainframes, for PCs, for the cloud, for every piece of infrastructure and technology that's ever been invented. So you know, the question of where the productivity is it's sitting in front of us but we have to implement it well and once we decide where to focus we have the issues of implementation quality of us offering training change management super manager roles, changing job titles, all of that messy stuff that has to happen. So when you read a story like Block and Jack Dorsey laying off 40% of the people believing he's suddenly getting great productivity that is just a very misleading situation. That's kind of my advice for today. We have many many examples of this that are going to be presented at Irresistible 2026. I hope you join us and we will be at Unleash in Vegas we will be at Transform in Vegas. I will be at the HR Tech conference in Europe and April and many many more places around the world if
[00:19:58] Speaker A: you'd like to talk to us.
[00:19:59] Speaker B: But come to Los Angeles June 8th through 10th and you will have a chance to be a part of our fifth annual research conference and really see this stuff in action. That's it for.
