Episode Transcript
[00:00:06] Speaker A: Hello everyone. Today I want to talk about Microsoft just had earnings yesterday of $61.9 billion for the quarter increasing 17%. Operating income increased by 23%.
[00:00:22] Speaker B: And the most significant of all is.
[00:00:26] Speaker A: That the Microsoft copilot is now in.
[00:00:29] Speaker B: Use, according to Microsoft, by 60% of the Fortune 500.
[00:00:35] Speaker A: So in a little more than a year, maybe a year and a half through Microsoft's relationship with OpenAI, the company.
[00:00:43] Speaker B: Has delivered by far the largest, most prevalent AI tool used in the world.
[00:00:51] Speaker A: And I just read a research report.
[00:00:54] Speaker B: By Comptia who studies it, that something like 94% of professional users, business people, have used some form of AI. In other words, chat, GPT, Microsoft, perhaps Google, perhaps anthropic, perhaps another.
[00:01:14] Speaker A: And as most of you know, the.
[00:01:16] Speaker B: Little AI bot is now included in WhatsApp, it's included in Facebook, it's going.
[00:01:23] Speaker A: To be included in Instagram. We also saw an announcement by Facebook meta that they are putting billions of dollars into AI. Actually, their earnings were a little bit down because of this. But to AI enable the user experience in our social media, obviously in our desktop applications and our corporate applications. By the way, the CompTIA study also showed that 61% of companies are already using AI in their corporate infrastructure for some applications.
[00:01:53] Speaker B: Now the reason Microsoft is important is.
[00:01:56] Speaker A: Because they do have technology that sits on the desktops. By the way, we use Microsoft here.
[00:02:03] Speaker B: And all of my Microsoft computers now have the copilot for free, not the.
[00:02:08] Speaker A: Enterprise version, but the personal version. So they are deploying AI at a.
[00:02:13] Speaker B: Scale that's just spectacular.
[00:02:17] Speaker A: Obviously it's costing them a fortune to run all the servers behind this, but.
[00:02:21] Speaker B: They'Re willing to tolerate that because the.
[00:02:24] Speaker A: Copilot revenue to Microsoft is going to be massive. And what this means, in my personal.
[00:02:29] Speaker B: Opinion, and we'll see where this goes, is that whatever you're building in your.
[00:02:35] Speaker A: Company, or whether you're buying it from workday or ServiceNow or Galileo or whatever it may be, you have to accommodate the fact that many, many, many of.
[00:02:46] Speaker B: Your users will probably have Microsoft products.
[00:02:51] Speaker A: And there really doesn't seem to be an alternative from Google or anybody else at the moment. OpenAI is certainly making a stab at the enterprise market, but they're not going to have a chance with this much.
[00:03:02] Speaker B: Microsoft technology already in place and they're just getting started.
[00:03:06] Speaker A: I mean, there's a big announcement of copilot stuff coming in May at the Microsoft Build conference, which I just heard about. And it's just going to go faster.
[00:03:15] Speaker B: And faster and faster now as you.
[00:03:18] Speaker A: Learn about how these AI systems work.
[00:03:21] Speaker B: There's layers here.
[00:03:22] Speaker A: The top layer, of course, is the conversational interface where you type or you talk to it, and it interprets that and turns it into various prompts and commands behind the scenes. But there's an orchestration layer that's going on under the covers. When you type a question or talk to it and tell it what to.
[00:03:43] Speaker B: Do and create a prompt, it's using the prompt to invoke a whole series.
[00:03:49] Speaker A: Of steps and activities that it has.
[00:03:51] Speaker B: Been trained to do well.
[00:03:53] Speaker A: You as a user or as a developer, can actually get under the covers and affect what that orchestration looks like. So if at some point in the prompt chain of asking questions, you want the copilot or chatbot to fire off an assessment.
[00:04:12] Speaker B: If you're a candidate or open a window in another application or send a.
[00:04:18] Speaker A: Message to somebody else, you can do that. So in some sense, from a Microsoft.
[00:04:23] Speaker B: Standpoint, the copilot isn't a standalone end.
[00:04:27] Speaker A: User thing like Microsoft Word. It's a tool. And the enterprise or corporate version of the Microsoft Copilot gives you all sorts.
[00:04:35] Speaker B: Of tuning knobs under the covers to.
[00:04:39] Speaker A: Change the orchestration, add things to it, add apps to it. There will be multiple copilots that you'll be able to create, that you'll be able to integrate with it, and other vendors who provide HR related solutions will basically coexist with it. Now, I spent several days with Paradox last week, and they are by far the leaders in conversational AI in all aspects of recruiting. And the paradox system is very, very sophisticated, is very easy to use, and very highly trained in many, many of the things that happen in recruiting and also in employee experience. You know, Microsoft's gonna impact them, too, to some degree. They're gonna have to figure out what their coexistence strategy is as well. So. But all I can say is if you look at the revenues and the financial growth of Microsoft, you can't deny.
[00:05:28] Speaker B: That this stuff is working. Now, is it improving productivity?
[00:05:32] Speaker A: Absolutely. Lots of studies are starting to come out, and there's more coming out about people saving hours a week and multiple hours per day. I've interviewed a lot of our Galileo customers, and frankly, most of them have told me they're saving almost an hour a day by using Galileo for many, many things, not just looking up information in Galileo, but using Galileo to access other information that they have inside of their companies.
[00:05:58] Speaker B: So these are in some sense the new Microsoft office.
[00:06:02] Speaker A: In fact, when I was talking to Microsoft this morning, the reaction that I had was when Microsoft office first came out, I don't remember when it was in the 1980s. I mean, we didn't have anything like this before.
[00:06:14] Speaker B: We might have had Visicalc or Wordplan.
[00:06:18] Speaker A: Where we had some of these funky.
[00:06:20] Speaker B: Microsoft based Windows tools for productivity.
[00:06:25] Speaker A: And then this suite of applications came out that did all these things from Microsoft, by the way. They had copied Lotus 123, who actually did this first. And all of a sudden, everybody's life changed. And here we are 25, 30 years later, and we're still using Microsoft office.
[00:06:40] Speaker B: Well, in some sense, the copilot is that it is Microsoft Office 3.0 or.
[00:06:46] Speaker A: 4.0 or whatever number you want to use.
[00:06:49] Speaker B: That is going to become our day to day activity center for many, many.
[00:06:56] Speaker A: Many of the things that we do as desk or office or professional workers. Now, those of you that are, you know, remote or you work in plants or offices or hospitals or retail stores, you're using your phones, but the same thing is happening on the phone. And it may be that Siri, in its new manifestations, takes on the role of the copilot for Apple. And there will probably be a Google assistant that will take on the role of the copilot for Android.
[00:07:26] Speaker B: But this is a giant change in.
[00:07:29] Speaker A: The platforms that we use to do.
[00:07:32] Speaker B: Our business work, to communicate, to learn, and to basically live our lives.
[00:07:39] Speaker A: Now, let me talk about some of the stuff under the covers. We have spent a lot of time on AI with you guys. We're going to be introducing a big paper on enterprise talent intelligence week after next. You're going to want to read it because it's very significant.
It's about the evolution of AI in the platform space. I talked briefly last week about workday.
[00:08:02] Speaker B: Let me talk about workday and SAP.
[00:08:05] Speaker A: What does this enterprise AI copilot stuff.
[00:08:09] Speaker B: Mean to all of the applications you buy?
[00:08:13] Speaker A: Well, first of all, as you probably.
[00:08:14] Speaker B: Know, we're going to be interacting with.
[00:08:17] Speaker A: These systems much more through chat and much less through buttons and panels and tabs and all the things we've had for years. So you could predict that five years.
[00:08:27] Speaker B: From now, the chatbot will be the.
[00:08:30] Speaker A: Application, and you will just ask it or tell it what you want to do, and it'll find the relevant information.
[00:08:36] Speaker B: It'll kick off the transaction.
[00:08:38] Speaker A: It'll give you what you need. We're not there yet. We're not going to be there for a long time, but that's going to happen.
[00:08:43] Speaker B: In the meantime, though, what the vendors are doing is integrating AI into their core applications.
[00:08:50] Speaker A: I mentioned last week that Workday described its plans to create micro llms that facilitate and improve the interaction of many, many aspects of workday. And that's great because there's lots and lots of complexity to workday and sometimes it's very hard to use. SAP is doing even more than that.
[00:09:07] Speaker B: SAP is building global llms to power.
[00:09:12] Speaker A: Success factors in all of the decisions get made across the whole suite.
I think successfactors in some sense is ahead of workday in this. When you're looking for a candidate, or creating a job description, or assessing someone, or deciding what role somebody fits in, or finishing a performance review and creating a development plan, successfactors can use its AI to assist you with all of those decisions based on the massive amount of human capital information in success factors as a whole. They also have standalone smaller LLM features, but they've implemented it in a more global way.
[00:09:50] Speaker B: They have also built Joule. Joule, J o U L e, is.
[00:09:55] Speaker A: The Microsoft copilot equivalent for SAP, and it is being trained, and has been.
[00:10:00] Speaker B: Trained to understand not only all of.
[00:10:03] Speaker A: The transactions and use cases and data reporting in successfactors, but in all the other SAP products. So if you're an SAP customer and you're running SAP financials and supply chain and HR, you're going to be able.
[00:10:17] Speaker B: To give a dual user interface to.
[00:10:20] Speaker A: All of your employees and they can ask it anything from, you know, how many days left do I have vacation to what was my expense account last month, and has it been paid yet? And what is my 401k balance? Whatever you want to ask it, and it'll find the relevant transaction in somewhere in the SAP infrastructure and provide it to you. That's something worked, hasn't done yet. I mean, they will eventually have to get around to it, but that's where SAP is going. So in some sense, I think SAP is maybe a year ahead in its manifestation of AI at the corporate space in the areas of recruiting and development and internal mobility.
[00:11:00] Speaker B: Products like Eightfold, gloat, Beamery, hired score, which was just acquired by workday Seekout phenom.
[00:11:10] Speaker A: Many, many of the ISIMs, many of the recruiting tools that we think of as applicant tracking systems or recruiting automation systems are now AI powered, and they are using what we call enterprise talent intelligence, which is external data, not just internal data, by the way. When you use the copilot, you're only looking at data that's yours. When you use Galileo, you're only looking at data that you're allowed to see. But when you use Eightfold or when you use, seek out. You're looking at data across the whole Internet, and that system is in some sense much more intelligent because of the massive amount of data that it has. And these talent intelligence systems, which started.
[00:11:52] Speaker B: In recruiting, have now become enterprise class systems.
[00:11:56] Speaker A: And that's really what the research is about that we're going to be launching is talent intelligence isn't just a way to source and select people for hiring. It's a way to identify two jobs in the company that might overlap with each other but are paid differently. It's a way to improve internal mobility and create a talent marketplace, which is what gloat specializes in.
[00:12:17] Speaker B: It's a way to identify pay inequities.
[00:12:20] Speaker A: When somebody with high skills might be underpaid against somebody with low skills, but they happen to have been around longer, so they've had more raises. It's a way to look at leadership candidates work that Heydrich and struggles is doing with Eightfold. To bring leadership assessment data together with.
[00:12:39] Speaker B: Skills data to better find hidden leaders.
[00:12:42] Speaker A: In your company is massively important. That's a groundbreaking solution. And the opportunity for the talent intelligence system to inform and improve performance management. Now, the sort of more primitive AI implementations of performance management today are making.
[00:12:58] Speaker B: It easier to write the appraisal using.
[00:13:01] Speaker A: Generative AI to write text. But the more advanced ones, which I've seen companies do in Eightfold and other platforms, is where you take actual performance data, sales data, number of lines of code written, or whatever you might be responsible for, and you use the AI.
[00:13:20] Speaker B: To inform not only feedback from peers.
[00:13:23] Speaker A: And what they have to say about your performance, but what you actually got.
[00:13:26] Speaker B: Done in a much more meaningful way.
[00:13:30] Speaker A: Now, I don't know that anybody's really using that yet for true performance management, but I do know one company, there's one software company that's doing that, and they just laid a bunch of people off and they are using that information to help them decide who they needed to let go. So this is where that's going. And that area of enterprise talent intelligence, which uses internal and external data, is really starting to encroach on many, many of the use cases in the HCM platforms. Certainly recruiting, certainly internal mobility, certainly career management, training, learning and development and so forth.
So just like the copilot is massive on the front end implementations of AI, including obviously chat, GBT, talent intelligence is doing the same thing on the back end.
[00:14:16] Speaker B: And I think we're going to have.
[00:14:18] Speaker A: One of the most disruptive periods in the HR tech market in my entire.
[00:14:22] Speaker B: Career, to be honest. The final category of this that I want to mention is corporate learning and development.
[00:14:30] Speaker A: And I've been spending a lot of time with training providers the last couple of months, and we're going to really dig into this and write some research reports on it later this year and early next year.
[00:14:40] Speaker B: The learning and development space is really old.
[00:14:45] Speaker A: I mean, it really pivoted in 1990 819 99 towards elearning.
Most of the learning systems we had before that were basically classroom management systems and scheduling systems. We had a good 25 years building learning management systems that could handle video content and scorm courses and simulations and.
[00:15:07] Speaker B: Tests and articles and audios, video and.
[00:15:12] Speaker A: All sorts of content, but in a very structured form or in the form of a portal, a learning experience plan, a learning experience platform portal that you could search and index and categorize by topic or by skill. But we really didn't have that sophisticated a world of content generation or content search, really. You could search for a course, but you couldn't search for much inside of a course.
[00:15:39] Speaker B: Well, forget it guys.
[00:15:41] Speaker A: It's all about to change in a huge way.
[00:15:45] Speaker B: As you know from what we've talked.
[00:15:47] Speaker A: About with Sana and other vendors, Dochebo is working on this, uplimit is working on this, other companies are working on this. You can take a document, you can take a video, you can take an article, you can take an old course, you can put it into the AI, and you can generate a new course.
[00:16:06] Speaker B: That'S 60, 70, 80% ready to go.
[00:16:09] Speaker A: The course can have chapters, it can have assessments, it can have topics, it can have graphic and media objects created automatically, and then you can go in.
[00:16:19] Speaker B: And edit that and you can edit it using AI.
[00:16:21] Speaker A: I just had a conversation this week with a large distributor in Europe who happens to be one of Sana's big customers.
[00:16:29] Speaker B: She told me that her learning and.
[00:16:32] Speaker A: Development department is not really that big, but because of the AI generation capabilities, there are 200 people not in l.
[00:16:40] Speaker B: And D building really important content for.
[00:16:43] Speaker A: Their sales and distribution channel without having to do formal instructional design and content development.
[00:16:50] Speaker B: This is just the beginning of that.
[00:16:52] Speaker A: It's going to get better and better and better. This product called Shape from Dogeboard can take videos and audios of sales presentations and build sales training in a very high fidelity format almost completely autonomously. And it's going to touch everything. Obviously, the social media companies are a little worried about how easy it is to build content in YouTube and Instagram. But for us in the corporate world, we'll be able to build things and tweak them and edit them in corporate training like we never had before. But I think the bigger part of.
[00:17:24] Speaker B: This is on the back end where.
[00:17:27] Speaker A: Learning management systems kind of ran out of gas. Learning management systems are very complex. RAT's nested process management systems for compliance, processing, tracking, certification, learning paths, assessment. I mean all these, you know, very gritty things that you have to do in various forms of training around the different parts of companies. And a lot of training is mandatory and a lot of training is compliance and safety oriented and you have to track it and you need to know if people completed it, you have to know if they learn anything and so forth.
[00:18:00] Speaker B: But there's also, I think, a hundred.
[00:18:03] Speaker A: Times more content that isn't in the.
[00:18:06] Speaker B: LMS, that belongs in there.
[00:18:08] Speaker A: Compliance content, process content. We just did an upgrade to ERP, we did an upgrade to Salesforce, we added a new feature to this, we did a new process for that. We changed the pricing scheduled for this thing. We have new services over here. Where do you put all that stuff? You can't build courses for all these things. Nobody would even take them and nobody has time to build all those courses. Well, this is the kind of information.
[00:18:32] Speaker B: That can be ingested into these next generation AI platforms literally, automagically.
[00:18:39] Speaker A: And we've seen this with Galileo. We have fed Galileo 25 years of research, hundreds of videos, podcasts, articles. Now we're going to be having some third party trusted content providers adding to us, and it's in all these different forms. And with a little bit of work, not a huge amount of work, we can make it completely consumable and searchable and intelligent in Galileo. So you can not only take a course on the new pricing model, you can ask a question about the new pricing model, and you can ask a question about the old pricing model and the new pricing model. And you can ask the AI to take this proposal that was done in the old pricing model last year and redo it in the new pricing model for this year. I mean, imagine that's just one tiny example of how many things like this we're going to be able to do. And that in the L and D world is a huge change.
I always felt somewhere in the back.
[00:19:34] Speaker B: Of my mind that someday knowledge management.
[00:19:38] Speaker A: And learning would come together. It never really did because knowledge management was really a searching problem, not really a learning problem, it was really an indexing problem. But I think those two things are going to come together now and the platform oriented AI solutions in L and D are going to explode. The value and utility of L and D. And I think the people that run L and D systems and build all these programs and chief learning officers and chief talent officers that manage these things are going to have to think differently about what we can do. And it is going to be deeply involved because a lot of the knowledge management systems that already exist in companies are locked up in sharepoint or confluence or a wiki somewhere. And we'll have to make a decision.
[00:20:24] Speaker B: Whether they stay there or they move.
[00:20:27] Speaker A: Into this new super lms, which I don't know what to call it yet, that will have these vast amounts of searching generation capability and indexing capability that we really never had before. Now, the interesting thing to me about AI is not just the data management and generation capability, but the fact that.
[00:20:50] Speaker B: It learns my personal experience with our.
[00:20:53] Speaker A: Product, with Galileo, which, by the way, we rolled it out this week to all of our clients. So if you're a corporate member, you have access to it, or you certainly will by next week and make sure you're getting in there and using it.
[00:21:06] Speaker B: Is that it gets smarter all the time.
[00:21:08] Speaker A: We loaded a whole bunch of benchmark.
[00:21:10] Speaker B: Data in there last week, and we've.
[00:21:12] Speaker A: Been playing around with it to make sure it was formatted correctly.
[00:21:15] Speaker B: And when I logged in today, Galileo asked me if I wanted to know whether my turnover rate was competitive with.
[00:21:24] Speaker A: The peers in my industry. It just asked me that.
[00:21:27] Speaker B: I didn't even prompt it.
[00:21:29] Speaker A: It's smart enough to know what I do and what kinds of questions I ask and who I am and what kinds of questions that I have asked.
[00:21:37] Speaker B: In the past and how its data corpus is changing. And it tells me and sort of prompts me and suggests to me things.
[00:21:46] Speaker A: That I might be interested in. We never had any other systems that did that. I mean, you had to program that stuff in the past. So not only is the AI easy to use as a desktop user and easy to use for all these enterprise.
[00:21:58] Speaker B: Applications, but it is much, much more.
[00:22:01] Speaker A: Intelligent from that standpoint than anything we've ever had before. Now, this is the week. We have three weeks left before irresistible. The conference is almost full. There are some seats left. We are going to have a panel of AI experts, including the head of AI at Microsoft, the head chairman of the EEOC, some folks from the EU, talking about legal regulations and safety issues in AI. At the conference, which is still out there. We're going to show you Galileo at the conference. We're going to show you the new features and the new data sets of Galileo that we're launching. We're going to have a panel talking about systemic HR, some of the greatest implementations of systemic HR I've ever seen. We're going to have a panel talking about advanced l and D, just like I talked about earlier, because we want you to hear what's going on in some really cool companies there. We have a whole workshop on the four day workweek. We have a whole workshop on employee activation. We have a workshop on internal mobility, workshop on talent intelligence. You got to come, you guys. USC will be out of session.
[00:23:06] Speaker B: There won't be any protesting going on.
[00:23:08] Speaker A: The campus is going to be beautiful. We're going to have a great time, and we are also doing some very special things for ChRO. So if you come as a Chro and you bring your team with you, you can join the Chro track and you're going to get some other special activities that are not available to the rest of the attendees. So that's my message, my little advertisement there. I guess my final kind of concluding.
[00:23:32] Speaker B: Thought as I think about Microsoft and.
[00:23:35] Speaker A: What came out this week in their earnings.
[00:23:37] Speaker B: Whenever there's a major technological shift, there's.
[00:23:41] Speaker A: A massive amount of positioning that goes on in the tech space where every vendor tries to figure out what they.
[00:23:48] Speaker B: Can do to position themselves into the.
[00:23:52] Speaker A: New architecture and the new world. Obviously, we know what Google's doing. We know what Microsoft's doing. We know what OpenAI is doing.
[00:23:59] Speaker B: We're going to learn more about what.
[00:24:01] Speaker A: ServiceNow is doing at their conference in a week and a half.
[00:24:06] Speaker B: But we don't know what Apple's doing yet.
[00:24:08] Speaker A: We don't know what Twitter's going to do. We don't really know what Tesla is.
[00:24:13] Speaker B: Going to do exactly. We kind of do, and many, many, many more companies.
[00:24:19] Speaker A: So as an analyst and for those of you that are interested in tech.
[00:24:22] Speaker B: And buy tech, we're going to see.
[00:24:24] Speaker A: A lot of changes and I think.
[00:24:27] Speaker B: The players could be disrupted. The companies that figure this out well and implement it in a meaningful way.
[00:24:35] Speaker A: Are going to be around for a long, long time. The companies that don't pay attention to this stuff and maybe just do small.
[00:24:42] Speaker B: Tactical implementations may find themselves in a.
[00:24:46] Speaker A: Very difficult situation a year or two from now. And this gets back to what I'm so excited about from Microsoft.
[00:24:53] Speaker B: Microsoft is an old company.
[00:24:55] Speaker A: It's not quite as old as me, but it's been around a long time. I remember when Microsoft was founded, and.
[00:25:01] Speaker B: They have proven that even as a.
[00:25:03] Speaker A: Large company, when you're willing to take risks and you're willing to place bets and you have a forgiving environment and you hire great people and you motivate them and you build a culture of growth. You can adapt much more quickly than you realize. When I first brought up the idea of our copilot and our company, nobody.
[00:25:24] Speaker B: Wanted to hear about it.
[00:25:26] Speaker A: But here we are 15 months later.
[00:25:29] Speaker B: With almost 2 million employees working in.
[00:25:33] Speaker A: Companies that have access to Galileo. I mean, that happened in a very short period of time, and only because we leaned into the technology and adapted.
[00:25:43] Speaker B: It as quickly as we could.
[00:25:45] Speaker A: Those of you in HR that feel like things are moving too slow, you have to speed up. Your employees are using this at home. They are bringing their home implementations of AI into your office, and they are then wondering when the company is going to have some of the capabilities that they have at home. You remember how that happened with mobile phones and that's happened with other technology in the past. We have to move just as quickly as the consumer market, in a sense to take advantage of the benefits of this technology. So anyway, that's a little bit about what's been going on in my mind this week, and I will tell you much more next week I'm going to be in Europe and I'll tell you what's going on over there and what we learn in the HR tech market in the European Community. Have a great week.
