Episode Transcript
[00:00:05] Speaker A: Good morning, everyone. Today I want to talk about increasing intelligence of AI and the specialization of AI and give you my perspectives on Galileo because we're releasing it now to all of our clients and we're getting.
[00:00:17] Speaker B: Incredible feedback and learning a lot.
[00:00:21] Speaker A: So the big transformation that's going on.
[00:00:24] Speaker B: In the Internet and basically in our entire lives is this shift from the old search and find paradigm of the.
[00:00:33] Speaker A: Internet, obviously interrupted by advertisements and search.
[00:00:36] Speaker B: Engines which are trying to get us to buy things, now replaced by ask.
[00:00:42] Speaker A: A question and get an answer.
[00:00:43] Speaker B: And that paradigm change is very, very significant.
[00:00:47] Speaker A: And I think we as human beings much prefer to ask a question and get an answer than to guess through.
[00:00:53] Speaker B: Some search criteria how to find something.
[00:00:56] Speaker A: And by the way, that includes whether you're going into workday, your expense reporting system, your finance system, or to buy something on the web or to just learn something. Clicking and searching and looking through panels.
[00:01:09] Speaker B: And scrolling is not a lot of fun.
[00:01:12] Speaker A: It was in the beginning of the Internet for sure. I mean, it was a lot better than those green screens I used at IBM. But those days have come to an end, and that's been a good 25 years of fun. So now we're going to be asking questions. And on the back end of that question answering experience, there's going to be an intelligent machine generating answers. It's going to generate an answer, not just give us an answer, and it's going to generate that answer from its knowledge, the system's knowledge. So if you go to a shopping system like Macy's or some other clothing site, and it has a generative AI, it will answer you what kinds of clothes to buy, what color, what brands, how it should fit, and so forth. And it'll be great. It'll be a lot easier than scrolling through dozens and dozens of products and.
[00:01:56] Speaker B: Trying to find the one you want. In the case of our system, it's.
[00:01:59] Speaker A: Already been very, very deeply trained on hundreds of topics, case studies, examples, vendors, stories, research, maturity models, benchmarks in all aspects of human capital, HR management, leadership. And so it is giving us and our clients very intelligent answers to questions. And the reason it does that is not only because it uses a great generative AI engine, but there's a lot of those, but it's very, very well trained. And as we've already seen in Galileo, now that it's been in use for a year, is it's getting smarter all the time. So as we add more data to it, and we're going to announce a whole bunch of new things at our conference in May, it's getting even more intelligent. And when you ask it a question, it suggests other questions to help you.
[00:02:47] Speaker B: Better dive into the topic you're trying.
[00:02:50] Speaker A: To discover and answer the question or solve the problem you're trying to address. Now, what is this thing really? We call it an information assistant, but it's more than that. It's a consultant. It's a problem solving tool, it's a learning tool, it's an education tool. It's a way to feel more confident. We had a long conversation at the transform conference with a lot of HR leaders about this idea of HR confidence or the imposter syndrome. When you're in a meeting and somebody starts bringing something up, then you're supposed to know a whole bunch about it, but you don't really know that much about it and you want to know what to say. You can overcome all of those things with it. Now in the broad scope of technologies in general, you know, since I've been around since the seventies, this is a huge change and we're going to be experiencing the breakdown of Google. I think Google's monopoly is over. I don't know about you guys, I don't find Google that useful anymore. As a search engine, all I get is ads. I actually would like to get some information that isn't an ad, but now I can't even tell whether the thing I'm reading is an ad at all. So I'm hardly trusting it. LlM from OpenAI is pretty good, but it's filled with information with no sources. Perplexity is better. We are going to look for more.
[00:04:00] Speaker B: Trusted sources of information.
[00:04:02] Speaker A: I can guarantee that. Now not everybody, but most of us are going to move away over time.
[00:04:08] Speaker B: From these search oriented, click scroll find.
[00:04:12] Speaker A: Oriented interfaces towards trusted agents. And I mean trusted in the sense.
[00:04:16] Speaker B: That if you're an exercise freak or.
[00:04:19] Speaker A: You have a chronic illness, or you love a certain type of food, or you're an HR geek like us, you're going to look for an agent that really knows what you're interested in and has a lot of trusted information. And that's not going to be Google. And I don't have anything against Google.
[00:04:34] Speaker B: My son works there.
[00:04:36] Speaker A: But I just think their day has come and gone and they're going to adapt and they're going to try to turn Gemini into something like this. But I think the future of AI is specialized applications and systems that are very, very good at something and we use them for that and then they will talk to each other. There's this idea of meetings of experts or network of experts, where the AI's talk to each other. And that's where I think this is going to go. These massively large language models that are.
[00:05:04] Speaker B: Trained on billions and billions and billions.
[00:05:06] Speaker A: Of articles and content from the web are probably never going to be that good at detailed, specific answers to a particular domain. And so while of course Microsoft and Google and perhaps Facebook want to build these as marketing and sales tools and revenue generating tools for advertising, I think our daily life is going to be relying on much more specific systems. And you know, it's an analogy to how the Internet evolved that we now go to our favorite websites, our favorite retailers, our favorite information providers. We don't go to what used to be called Yahoo to find everything. We go to a search engine, but the search engine just points us in another direction.
[00:05:46] Speaker B: So that's where to me this is going.
[00:05:48] Speaker A: Now in the case of Galileo, which you can get as a corporate member now, so it's ready. We've discovered a lot of things. Not only only are these systems extremely surprisingly insightful, they give you very important information that you may not have found.
[00:06:03] Speaker B: Otherwise, because even if you do search.
[00:06:05] Speaker A: Around and click and read, you're not going to read everything on a topic. You don't have time. But Galileo does that and it selects what you want. It shows you the references and you can dynamically look in the reference at the location of the content that was referred and find it without having to read page after page after page to.
[00:06:23] Speaker B: Find what you need.
[00:06:24] Speaker A: We've recently started loading trusted, authoritative third party content and we're going to be building and launching a content partnership program. We have two large content providers we're working with right now, one for benchmarking information in HR and another one for global regulatory compliance and employment laws around the world. And we'll be talking more about those relationships at the conference.
[00:06:45] Speaker B: On the benchmarking one, I was really.
[00:06:47] Speaker A: Surprised what happened when we started loading the data in there. I wasn't really sure how it's going to respond. Systems started to ask me questions about my operation and whether and why my operation would be within or outside of these benchmarks. Extremely insightful.
[00:07:03] Speaker B: Very quickly.
[00:07:04] Speaker A: Now, it didn't do that just because it's an LLM. It did that because it already knows a lot about HR and business and leadership and management and all these employee experience issues. So this is sort of a big.
[00:07:15] Speaker B: Change coming now I'm going to be.
[00:07:17] Speaker A: Talking at the unleashed conference and then at the HR Tech conference in Europe, and then of course throughout the year and at our conference in May about where all this is going. And it's very clear to me that in the case of HR tech workday, Oracle, SAP and all the applicant tracking systems, learning management system and other transactional.
[00:07:35] Speaker B: Systems, we're going to be having a.
[00:07:37] Speaker A: New generation of software. At this point, the most descriptive term for it is what I call talent intelligence. And we're going to be releasing a big paper in a couple weeks on what we call enterprise talent intelligence. And what you'll see is that just like Galileo is a guru on HR, and by the way, it will become a guru on your HR issues in.
[00:07:58] Speaker B: Your company as you train it and.
[00:08:00] Speaker A: Load your information into it. The whole HR tech market is going.
[00:08:04] Speaker B: To move in this direction.
[00:08:05] Speaker A: If you have an ERP or an HCM system today and you're trying to ask us some questions about turnover or performance or productivity or, you know, do, do a leadership assessment of who our top leaders are or the financial impact of this reorg or span of control.
[00:08:22] Speaker B: That'S a lot of work.
[00:08:24] Speaker A: It's really hard to do it. The data is not very good. Most of the data is in multiple systems. And so those kinds of projects get.
[00:08:30] Speaker B: Delegated to an analytics group who builds.
[00:08:33] Speaker A: Their own data set and runs those kinds of reports.
[00:08:36] Speaker B: By the way, we're doing a big.
[00:08:37] Speaker A: Study of people analytics, so please participate in that if you'd like to get the results. But that's kind of why we don't do that kind of analysis very much. It's too hard. Well, imagine when that can be done almost immediately with a talent intelligence platform that has all the information about all your workers enriched with detailed information about their job history and experience and skills that you probably don't have in your HCM. You're going to be able to run.
[00:09:01] Speaker B: Those kinds of analyses all the time.
[00:09:04] Speaker A: And then you're going to be able to build AI models, to find future leaders, to identify patterns of low productivity, to find potential fraud, etcetera, in the HCM system, in the talent intelligence system. And that's what enterprise and talent intelligence is about. And I don't think the big HCM vendors understand this yet, because they have a lot of transactional applications they've built that are not designed this way. And there's more to come. One of the things that Galileo can also do is it can analyze tabular data. So as we load benchmark data and other forms of generic data that cross industries, you are going to be able to go in there and ask data related questions without having to build a pivot table or some sort of fancy report to get that information. So obviously that'll be really helpful in accessing all the data we have but you'll be able to use this for your own data. And so the implications of these technologies are massive on the people, analytics groups, all the data reporting and analytics technologies and tools that we have, the data warehousing tools we have, etcetera. All of that is going to be, you know, significantly impacted by this and I know that the data management companies are all investing in AI heavily so there's a lot going on. It's very, very positive, all good stuff and very transformational. So I just want to give you some things to think about and I really hope to see everybody in LA, we're about two thirds full of irresistible now and it will fill up. So if you'd like to come, you know, get yourself a seat, bring your team and we'll see you there. Bye.
