Ep 47: Workforce Transformation via Responsible AI, Skills, and Total Talent Intelligence with Remy Glaisner

[00:00:09] Welcome, or welcome back to another episode of Elevate Your AIQ. In this session, I reconnect with my super smart friend, Remy Glaisner, a respected voice in talent intelligence, market research and HR strategies, and a fellow member of the Talent Intelligence Collective community. Remy shares insights from his multifaceted background in areas such as automation and HR, exploring the rise of skills-based organizations and the critical role of human skills in shaping modern teams.

[00:00:35] We talk about the transformative impact of AI agents on HR and workforce planning. We touch on AI governance and the importance of data integrity, and we discuss the importance of fostering a culture of continuous learning and collaboration. In the Talent Intelligence community, we often talk about how organizations can harness total talent intelligence to adapt to market shifts, and Remy and I certainly hit on this, and what it means from both data and organizational perspectives. I always have engaging and insightful conversations with Remy, so I know you'll appreciate that we finally recorded one. Enjoy it, and thanks for having me.

[00:01:05] Thanks for listening.

[00:01:07] Hi, everyone. Welcome to another episode of Elevate Your AIQ. I'm your host, Bob Pulver, and this morning, I'm catching up with my friend, Remy Blazner. How are you doing today, Remy?

[00:01:19] I'm doing fantastic, Bob. Thanks for inviting me.

[00:01:21] Absolutely. Thanks for joining me. We finally get to record one of our hopefully insightful conversations. Why don't we kick things off with you just giving folks a little bit about your background. You've done a lot of really interesting

[00:01:35] work over your career. I'd love to hear about it. Sure, yeah, no problem. So I'm an engineer by training, automation engineer by training. It was a long time ago now. Never really practiced that, but I always kept, you know, that kind of like technical kind of spirits inside of me. So I'll still play with that. But very, very soon in my career, I jumped into the business side of things. I'm an ex-gartner. Spent my seven, eight years there. You know, I did my duty. Service duty.

[00:02:03] Yeah. I was also with IDC for a while, research director. I was running a little consultancy business covering like robotics, automation, intelligence, things. So pretty familiar with the old sort of concept around AI, although nothing's really like gen AI that just popped up recently.

[00:02:22] And most recently about like, yeah, three, four years ago, I joined AWS. Didn't stay very, very long there. That's where I would say my career shifted from, you know, anything automation, robotics, anything in the system to something more HR.

[00:02:40] So as a matter of fact, when I joined AWS, I was part of HR for the first time in my life.

[00:02:47] So I would say I described myself as I fell in love with a lot of things like talent intelligence, you know, like the use of AI and algorithmic technologies in different ways.

[00:03:00] And I'm kind of attached to that, but that's what I cannot say level, but 10 or 20 years of experience with that.

[00:03:07] And most recently, like the past three years, two years ago, actually been consulting with various companies, helping people, playing your advisor here and there.

[00:03:17] Some companies are, you know, solution providers in the talent space at launch.

[00:03:23] So like playing a little bit with them and how do you do marketing?

[00:03:26] How do you design product?

[00:03:28] How do you do this and that?

[00:03:29] And sometimes also on the user side, which is actually mostly what I've been doing, anything related to like, hey, let's be a skills-based company or something, or organization, something like that.

[00:03:41] What do we do with that?

[00:03:42] You know, just frankly, I've been thinking.

[00:03:44] In a nutshell, that's me.

[00:03:46] So when you were at these analyst firms, you weren't necessarily focused on even HR technology at all, even though you weren't.

[00:03:53] To that point, I would say somehow a lot of those technologies were very similar and used in a similar way in that that was corporations.

[00:04:03] Whether you use that on the shop floor for manufacturing or you are trying to do something in HR, you're trying to be more efficient.

[00:04:10] How you use that, maybe the endpoint technology might be different.

[00:04:16] But frankly, the goals you have, the way you want to serve a business and business and everything, they are strikingly exactly the same.

[00:04:23] Yeah, I think that mirrors a lot of the perspective that I've taken when I try to sort of frame some of my prior assignments.

[00:04:34] I mean, even when I was at IBM for over two decades, I mean, I did a lot of very different things.

[00:04:40] It wasn't like, I think in the past people have looked at careers.

[00:04:45] Oh, you spent all this time at one particular organization.

[00:04:48] Didn't you get bored?

[00:04:50] Did you feel stagnant or whatever?

[00:04:52] It's like, well, no, because at a company like that, it's so big and there's so many different divisions and lines of business and areas that you can focus.

[00:05:02] And you can take some of the skills and knowledge that you have acquired for each of those and repurpose that in these other contexts.

[00:05:12] So it sounds like you were able to do that successfully as well.

[00:05:16] That's exactly that.

[00:05:17] I mean, I think that's what I'm trying to do.

[00:05:19] I think that's what I'm trying to do with like, try to bring a fair and maybe a strong sense of business into like the HR, I would say, way of thinking.

[00:05:32] Not that the HR way of thinking is wrong at all, but just like, how do you, when you say you want to align to the business, what does that mean exactly?

[00:05:41] How do you do that?

[00:05:42] Those are topics I'm working on.

[00:05:44] Like I was thinking about skill-based organization, one of the things I'm doing is not how do we hire better, but how invest in skill-based organization to help, you know, not just align it, but like make the organization more competitive.

[00:06:00] We'll definitely dive into this skills-based organization topic, but it's nice to hear someone else's experience where you've really gotten a sense of how this business is.

[00:06:14] One of the pieces of advice that I got from an old mentor at IBM was know your business, don't just preside over it.

[00:06:24] Part of that is, you know, getting that exposure in these different areas.

[00:06:28] And so in a single role, maybe you get that with like a, like a chief of staff kind of role, which, which I've held a couple of times, but otherwise you've got to always be looking for opportunities.

[00:06:43] If, if, if, if, if, if, if, if the organization doesn't explicitly sort of encourage that, I mean, that's more of a recent phenomenon where we've seen advancements in internal mobility and things like that.

[00:06:55] But I think to your point, you have all these people who are, who are loyal and they've been developing skills.

[00:07:01] They have the tacit, you know, knowledge and the domain knowledge and they understand what your company is trying to achieve.

[00:07:08] So why not put more, a lot more effort into that and your existing workforce and increasing what seems to be a trend of, of decreasing 10 years.

[00:07:19] Why not try to reverse that trend and keep people, you know, engaged and happy and showing them that you're investing in their career,

[00:07:29] as opposed to spending, you know, 6x or 8x or whatever it is to, to find that new, that new people.

[00:07:37] Well, first I don't really define it.

[00:07:39] I think, you know, and that's by the way, something that, you know, I see a lot with all my HR friends, like, you know, very, very soon when something new pops up, people try to define things.

[00:07:51] And I think it's good, but it's also like not so good.

[00:07:54] It's good because it allows people to kind of talk about the same thing in a way,

[00:07:58] but it's also like it limits a lot what it is.

[00:08:02] If I say, okay, skill-based organization, it's about me using skill to define how the workforce can work and, you know, how we will plan it and how we will hire and anything.

[00:08:11] But then I limit a skill-based organization to something very HR-like.

[00:08:16] And to me, it's not so much like what is or how do I define the ASP, you know,

[00:08:20] but the point where, you know, the business and HR actually start to communicate differently.

[00:08:26] I start in business differently.

[00:08:29] The way I usually talk about it is HR tend to see skills like money on a bank account or currency on a bank account.

[00:08:37] Different currencies, you know, you want, I mean, it's very like, I call that 2D.

[00:08:41] So you've got currencies, you've got like how many of them do you have and all you know.

[00:08:46] The business, on the other side, sees skills at the transaction.

[00:08:50] On the sense that, you know, I take those skills, when I put them together into a transaction, it creates somebody.

[00:08:56] And that's how the business works.

[00:08:58] And somehow skills, when you do that, it's just the exact same coin.

[00:09:03] You're just looking at two different faces of that coin.

[00:09:06] And it's both like you, both sides will say, yeah, yeah, I understand that.

[00:09:11] But actually making it work is something more complicated.

[00:09:14] And I think to me, the SBO, an SBO is a company, an organization that has understood how to look at both sides of the coin at the same time and take advantage of it.

[00:09:24] My understanding of how HR has approached this has been some reluctance to take on some of the heavy lifting, I'll call it.

[00:09:36] And so I wonder how progressive organizations are like communicating the value of what you just described and how to sort of make that shift, right?

[00:09:48] Because there will be naysayers that say, oh, here we go again with skills.

[00:09:53] Haven't we always been hiring, you know, for skills and competencies or whatever?

[00:09:59] Skills is not something, it's not something at stake.

[00:10:02] So like when an actual organization has been falling into kind of a, I won't say that trap, but kind of trying to map out all the skills we have everywhere in a very, very precise way.

[00:10:14] It's good and it's also not so useful.

[00:10:17] It's so good because the moment you're done with that, from when you started, it might have changed.

[00:10:24] It's not exactly the same.

[00:10:26] And it's also great, but what do you do exactly with that?

[00:10:30] So do you need to define everything?

[00:10:33] Do you need to map out your entire organization?

[00:10:36] Do you need to have like such a level of granularity when you do this?

[00:10:40] You know, there's something that you might not actually need right away.

[00:10:45] And I'm thinking about anything that is kind of like process-like.

[00:10:50] Say, for example, well, you know, when we say an organization is in manufacturing and you have to build a new factory.

[00:10:57] Whenever you build a new factory to the same product, you know what you will have.

[00:11:01] You will need so many people that will have so many skills.

[00:11:04] So basically, if the business decides, yes, we need a new factory, you don't necessarily need to like map out all the skills and everything.

[00:11:11] You already know how to do that.

[00:11:13] What you need to map out is when you are actually facing something where, you know, it doesn't, kind of like a usual process doesn't apply.

[00:11:23] When something, you know, that the market creates a variable that you don't know.

[00:11:28] What do we need to have?

[00:11:29] What do we need to get?

[00:11:30] What do we exactly have in bank and everything?

[00:11:33] And those actually are usually only 10 or 20% of the entire organization.

[00:11:37] And they are very specific.

[00:11:39] But like understanding what will actually correspond to something the business really know how to do, to something the business will have to respond as fast as possible.

[00:11:48] This is where it starts.

[00:11:50] And this is probably the value.

[00:11:51] And the value is, if I reframe that, well, when you will have to respond to that, thanks to our small bit of organization, you will not only be able to anticipate it, but to create scenarios around that.

[00:12:04] So to play around with it, like, oh, if I do that, what will happen?

[00:12:07] We'll help you from a talent perspective to understand what the implications are and help you just make the decision even before that.

[00:12:15] That's great value for the business, you know, because if something happens, market dynamics, something happens.

[00:12:22] You can commit whatever million dollars on hiring whoever.

[00:12:26] If you're late in the game, you're late in the game.

[00:12:28] You're not as competitive as the next door competitors.

[00:12:31] But if you can actually anticipate all that, then you win.

[00:12:34] And again, it doesn't necessarily take to all the skills to an organization and not so precisely.

[00:12:40] It just means that you understand how the business works and what the business might face as something that is unexpected in the sense that not following usual process or usual way the organization works.

[00:12:54] I absolutely appreciate everything that you're describing.

[00:12:58] I guess I just view it as one of the key components of your sort of agile strategy, right?

[00:13:06] Like, how will you be able to anticipate those market shifts, to your point?

[00:13:13] How do you best respond to them?

[00:13:16] Yeah.

[00:13:16] How do you enter a new market without, you know, proactively understanding what you'll need and where, you know, a gap may exist?

[00:13:27] Or how do you shift from using your example with manufacturing?

[00:13:30] If you're an automotive manufacturer and all of a sudden you decide to shift to, you know, electric vehicles, what are you going to do?

[00:13:39] Yeah, you might set up a new factory and assembly line and all of that.

[00:13:43] But you're obviously going to have some degree of skills, you know, gap that you need to understand.

[00:13:49] And now you need to look and say, okay, well, what do I know about transferable skills from these roles on a typical, you know, internal combustion, you know, assembly line versus whatever?

[00:14:03] There's definitely a lot of overlap.

[00:14:05] A car is already, even an internal combustion engine car is already sort of a computer on wheels to some degree, right?

[00:14:13] So there's definitely some pieces of that that are overlapping.

[00:14:17] But yeah, there's going to be net new skills that you need to acquire and make those adjustments.

[00:14:21] So, I mean, I think this ties into a lot of what you and I discussed with the community and talent intelligence collective, which is how do you take all of this, you know, skills data and the people analytics data and the talent intelligence, which may be looking outward versus people analytics looking inward.

[00:14:40] How do you take all of that and do proper sort of scenario planning?

[00:14:45] And to your earlier point, you can't just take all this time to do like this skills inventory at a granular level, because by the time you're done, things have changed.

[00:14:56] And I would say like, it's not that the old inventory is necessarily bad.

[00:15:00] It's just like, don't start with that.

[00:15:01] That would be my recommendation.

[00:15:03] Also, you know, I think the example, like the illustration you were giving me that the manufacturer is great.

[00:15:10] You know, I was talking about scenarios before.

[00:15:12] That's exactly that.

[00:15:13] That's like scenario one, we have some of the skills to move that.

[00:15:17] So do we upskill?

[00:15:18] Do we get new people?

[00:15:19] Scenario two, we just get new people.

[00:15:21] Scenario three could be like another combination of things.

[00:15:24] But that's, you know, something that HR will be able to offer and, you know, very detailed pros and cons on doing that.

[00:15:31] But doing it like it's, you know, before it happens in the sense that the business might be thinking and scratching their beard, like, hey, should we do that?

[00:15:39] Well, HR means you can say, well, from a time perspective, here's what will happen.

[00:15:44] Here is how it will impact our competitiveness in the market.

[00:15:49] Here is how fast we can do that.

[00:15:51] Here is how realistic that is.

[00:15:53] And that the business needs to know.

[00:15:54] Yeah, for sure.

[00:15:56] I guess the other thing I wanted to ask you about was the skills exercises where you talk about like skills intelligence and skills gap analysis and all of that.

[00:16:07] Do you see that focusing enough on like what we used to call soft skills?

[00:16:13] Skills?

[00:16:13] Yes.

[00:16:14] Most of the discussions that I've seen and participated in, they seem to be focusing on the hard, you know, technical, more tangible skills.

[00:16:23] Validating those skills can still be a challenge.

[00:16:26] So there's still some potentially some subjectivity there, but there's more subjectivity perhaps on the human skills because there's just so many ways to try to how many different types of assessments and whatever.

[00:16:40] And everyone thinks it's, you know, snake oil or it's not grounded in science or academia or whatever.

[00:16:47] So I guess I just wonder from your perspective, being much deeper into the skill space these days than me, like how you think about that?

[00:16:55] That's a good question.

[00:16:56] I would say right now was happening.

[00:16:58] People are more focusing on like hard skills, but usually pretty technical skills, like something that can be indeed measured or assessed or something like that.

[00:17:08] Now, do you like the soft skill have to be taken into consideration?

[00:17:12] I think yes.

[00:17:13] But let me backtrack just one second here.

[00:17:16] When you talk about skills, and that's actually one of the complications that come with skills is you can talk about skills at the individual level.

[00:17:25] You mean.

[00:17:26] So then in that case, yes, talking about, you know, soft skills on that, you know, communicate, do whatever.

[00:17:32] It's very important.

[00:17:33] You can also talk about skills at team level, at business unit level, as something like that.

[00:17:40] But then in those cases, maybe some form of soft skills will actually surface, but not always.

[00:17:50] It's more like, again, it's, you know, if soft skills enable faster transaction of skills for the business, then soft skills are very important.

[00:18:01] If soft skills are not helping or not changing fundamentally on those transactions, skill transaction existing, then, you know, it's nice to have.

[00:18:11] And that's also, by the way, you prioritize, should I do that or not?

[00:18:15] And linking that, by the way, to AI.

[00:18:17] I think AI is probably like the way to potentially start to integrate that and to see, well, how into the mix does it really play or not?

[00:18:25] I don't have an answer to a question, should we do that or not?

[00:18:27] But if we could either assess or, I don't like assessing this really, like, you know, observe how some individuals are having like XYZ soft skills on top of like hard skills.

[00:18:42] How do they play together?

[00:18:44] How do we record that?

[00:18:45] And now with AI, we can manipulate a lot more data, more information about like the skills, a lot more defining part of skills.

[00:18:55] Before we move on, I need to let you know about my friend Mark Pfeffer and his show, People Tech.

[00:19:01] If you're looking for the latest on product development, marketing, funding, big deals happening in talent acquisition, HR, HCM, that's the show you need to listen to.

[00:19:13] Go to the Work Defined Network, search up People Tech, Mark Pfeffer, you can find them anywhere.

[00:19:22] Yes, I mean, we might discover that in some cases it's very important.

[00:19:25] And we know that.

[00:19:26] I mean, people cannot communicate very well.

[00:19:29] I mean, in one way, they might say verbally, they might do very well in a written part.

[00:19:34] And that might be great on some occasion.

[00:19:37] That might not so great on other occasion.

[00:19:38] How do we play with that?

[00:19:40] So that might be, it has an importance.

[00:19:42] Is it being done?

[00:19:44] Probably not.

[00:19:45] Or not so, so much.

[00:19:48] But it should be.

[00:19:49] I guess it depends on the nature of the work and the industry.

[00:19:54] And you've got to look at the balance of efficiency versus effectiveness or whether you're really prioritizing the experience of particular persona versus just, again, trying to put everything under like a productivity kind of lens.

[00:20:14] Right.

[00:20:14] Not everything should be under productivity or efficiency.

[00:20:17] And that's actually where maybe like something more HR-like for the business.

[00:20:22] If you, thanks to skills, you know, based, whatever new framework you use, you start creating teams made of people who won't go along together at all.

[00:20:33] Who potentially will make that super critical team explode in six months.

[00:20:38] Well, you should actually be able to prevent that.

[00:20:41] Or to tell the business, yes, if we do that, you know, yeah, or we could have like BU1 and BU2 kind of like internally compete about that.

[00:20:48] And people murdering each other.

[00:20:49] But that might not work for that reason.

[00:20:51] Like, you know, we will get like, I don't know, we'll move from 20% of people like, you know, just leaving the company to 40% every year.

[00:20:58] Something like that.

[00:20:59] That's crazy.

[00:20:59] We cannot do that.

[00:21:00] But the business will listen to that if you can, I would say, like, you know, bring data about it.

[00:21:06] So, yes, it's very important.

[00:21:08] But it's just like, why?

[00:21:11] I mean, you just have to find a kind of like the use case or I would say the business case to which it creates an importance.

[00:21:18] It creates like, it decreases or increases the value of what you generate.

[00:21:22] That makes sense.

[00:21:23] I know we've talked about this and there have been some debates in the talent intelligence collective that you and I have both participated in.

[00:21:31] But when we think about like this concept of like total talent intelligence, right?

[00:21:38] Bringing together all of this data, wherever the people exist, wherever the skills exist, understanding market dynamics, understanding, you know, skills gaps and all of this stuff.

[00:21:48] And pulling together everything that we just talked about in a holistic way.

[00:21:53] Is it your observation that there's some progression towards people realizing how important it is to pull all these things together, breaking down some of the silos where a lot of this sort of people related data exists?

[00:22:09] Or there's still like, you know, territorial, you know, fights going on and this is going to be a perpetual battle?

[00:22:15] Yeah, I think it's a good question.

[00:22:18] I think there will always be some from a battle somewhere.

[00:22:21] You know, like my data is not your data, blah, blah, blah, blah, blah, blah, blah.

[00:22:24] But I think, you know, beyond the concept of tool, talent intelligence, I think it's important to think about the intelligence enterprise organization.

[00:22:34] So that's what it is.

[00:22:36] And the tool of talent intelligence, whatever you want to call that, say like the HR part of that intelligence of organization needs to exist and need to frame itself in a way that will actually integrate well the intelligence enterprise.

[00:22:52] Because that's what matters at the end of the day.

[00:22:54] You want your organization to be as intelligent as possible.

[00:22:58] One side of that is the HR side.

[00:23:00] I want another side of that is the marketing intelligence.

[00:23:03] Another side of that is, you know, finance intelligence.

[00:23:06] Another side of that.

[00:23:06] So, yes, it has to be done somehow.

[00:23:09] Now, I know that, you know, the talent intelligence collective, usually you start bringing a topic and people are set off.

[00:23:17] They are all over the place.

[00:23:18] It will never work.

[00:23:19] Sure.

[00:23:20] Yes.

[00:23:20] But if we start to like it will never work, then it's true in conversation.

[00:23:24] And I think it can work.

[00:23:27] It should work.

[00:23:27] But again, not necessarily as in just talent for talent intelligence and tool talent intelligence, but tool talent intelligence in the context of the intelligent organization.

[00:23:41] So what comes first?

[00:23:42] What comes second?

[00:23:43] How should those data be formed, be structured?

[00:23:46] So it actually helps us at, you know, more micro level, you know, like the HR level or maybe like any HR use case level, but also the organization.

[00:23:56] And that is like, you know, the organization part, the more like broad part, like, you know, how do we help, you know, finance be better at finance because of our talent, tool talent intelligence, blah, blah, blah.

[00:24:06] That I think is not necessarily true enough.

[00:24:08] And I think that I believe that a lot of reluctance right now also come from, you know, if I want to be the boss, you know, hey, TI should be the boss and control everything.

[00:24:20] You know, hey, it's a strategic purpose planning.

[00:24:22] Yes, I should be, we should be on top of it.

[00:24:24] Oh, no, no, no.

[00:24:24] So wait, wait, wait, wait.

[00:24:25] It's people, people are in it that should do everything.

[00:24:28] I think it's just about that.

[00:24:29] It's just like, yeah, it's very hard to exist when you are in HR.

[00:24:33] HR is like, it's a backup thing.

[00:24:35] It's not always spoiled very well by the business, budget unlimited and on and on and on.

[00:24:40] You want to be visible.

[00:24:41] You want to be there.

[00:24:42] You want to control everything.

[00:24:43] So I do understand that, you know, the organizational frictions.

[00:24:47] At the end of the day, totally total intelligence, totally total workforce planning strategy.

[00:24:53] Give it the name you want.

[00:24:55] At the end of the day, what matters is that it supports the intelligent enterprise.

[00:25:00] If it doesn't, if it just supports the intelligent HR for HR, then you fail and it won't work.

[00:25:06] Absolutely.

[00:25:06] I think there's similar sort of ownership conversations happening when it comes to responsible AI, actually, which I know we'll talk a little bit about later.

[00:25:17] But one of the things that I was thinking about as I spend more time learning about like agentic AI, you know, creating these AI agents, which are actually like, it's not just robotic process automation.

[00:25:33] And it's not just an AI sort of assistant or co-pilot.

[00:25:37] It's not just giving you answers to a question or whatever.

[00:25:41] It's actually executing something.

[00:25:43] And there's logic and there's reason.

[00:25:45] Well, we can debate separately whether it's actually doing its own reasoning or not.

[00:25:50] But the point is, it's executing some of these tasks that even two years and two weeks ago, before ChatGPT came out, we would have thought that these things were sort of the domain of human beings.

[00:26:02] And so when I think about some of the friction points and with appropriate controls and privacy and whatever around the data elements.

[00:26:12] So when you can start to sort of embed that and codify that into specific AI agents, and then you set up, you know, the workflows that actually connect those agents and orchestrate the data and the decisions or whatever.

[00:26:29] And you create this agentic AI sort of environment.

[00:26:32] One of the use cases that I think about where this would come in handy for HR and beyond is in this sort of total talent intelligence where you don't have to go and tap somebody in another department to create a net new dashboard.

[00:26:49] Or you don't need to go and find your data scientist or whatever to just give you a specific answer to a challenge that you're facing or whatever.

[00:27:02] You've got more of, you've sort of codified a bunch of expertise into one of my prior guests called it a digital advisor.

[00:27:10] But you don't have to actually go and worry about the friction and the scheduling time with this person and da-da-da-da.

[00:27:19] So I just wonder, you know, if you've explored some of these agentic concepts at all and how that might apply to this domain.

[00:27:28] Very honestly, I haven't been like hands-on so much with agent AI.

[00:27:33] But when you think about it, that's not more than the elaborate version of an application.

[00:27:40] So, you know, if you think about, I mean, and right now HR and the other like a major function, they already work with a bunch of applications, you know, that even on the cloud or not, most of them on the cloud.

[00:27:50] Let's use data, XYZ coming from XYZ platform or like data lake or anything.

[00:27:55] Just like it's another type of maybe interface to play with.

[00:28:01] To me, it's not just what it is.

[00:28:03] But that's mostly what it is.

[00:28:06] So how can it potentially limit friction?

[00:28:11] Well, yeah, it will be friction.

[00:28:12] I mean, fully make reluctant from, I don't know, if you were talking to someone, yeah, yeah, I'll do a dashboard later.

[00:28:18] Don't worry.

[00:28:19] I'm just reluctant to do it.

[00:28:20] So I'm just, it will take me two months.

[00:28:22] I will explain to you how complicated it is to not do it.

[00:28:25] Okay.

[00:28:25] You won't have that problem anymore, but you will have other problems.

[00:28:28] And it's more about, I would say, automating to a level that is like more fluid, a certain function, a certain activity, a certain like it's about doing things when generally AI is about creating things somehow.

[00:28:43] So, you know, just define a box and like that box will be a major agent.

[00:28:48] We all use agent AI.

[00:28:49] So we'll be able to like perform some complex tasks around XYZ to define.

[00:28:55] How is it so different from, I think, creating some form of self-serve dashboard that people can use?

[00:29:04] Well, maybe it will, you know, just do the job a little better, a little faster, make it a little more fancy.

[00:29:10] It won't just be a dashboard.

[00:29:11] They may be more visualization, maybe for other way, but this is just a layer of practically added to what already exists.

[00:29:22] So the friction, they will, you suppress them here.

[00:29:26] They will start to exist there.

[00:29:27] You know what I mean?

[00:29:28] And it's not necessarily like a good or bad.

[00:29:31] Like friction are, you know, when frictions happen, I think they, it's not necessarily negative.

[00:29:37] And as a matter of fact, it's often a positive thing that two different people, functions, have different perspectives into one thing.

[00:29:45] And it's not who has the right one.

[00:29:47] And when you see friction, it's about like, hey, tell me about how you see the world.

[00:29:52] I'll tell you how I see the world.

[00:29:53] Can we do something together?

[00:29:55] Now, as I said before in HR, very often, not always, but very often, you know, people are competing for a little bit of sunshine, I would say.

[00:30:06] So that creates like, you know, let me be more right than you are.

[00:30:10] I think there's a lot of willingness with those kind of pressures that exist.

[00:30:15] There's a lot of willingness in HR, and that's probably way more than any other usual corporate function to collaborate.

[00:30:22] But this is where the functions are coming from.

[00:30:24] So now, identity, yes, to your question, they will remove some friction, but only to create some other elsewhere.

[00:30:32] Yeah, it's not that any particular agent will necessarily have, will be like 100% correct, right?

[00:30:41] We're still not at the, you know, AI as a calculator, let's trust it implicitly kind of thing.

[00:30:46] It's still sort of codified representation of probably could be one person's opinion as sort of a, I don't know if digital twin is the right way to think about that.

[00:30:57] But yeah, that's also what could be.

[00:30:59] I mean, hopefully it's incorporated, you know, multiple perspectives and, you know, a whole separate conversation about whether, you know, an AI agent can actually encapsulate, self sort of encapsulate the collective intelligence of a group or a team or whatever.

[00:31:15] But the point is, it's been trained specifically to do a particular function like an appliance sort of, but give you in natural language, you know, some answer.

[00:31:26] But, you know, when it starts communicating that information on your behalf to perhaps another agent or to someone in another department or whatever.

[00:31:37] I mean, there's all kinds of controls that we need to think about.

[00:31:40] And when I put my Responsible AI, you know, hat on, I have to think about how we're actually sort of monitoring those types of things.

[00:31:49] But it's also an example of when you think about AI governance and assurance and stuff like that.

[00:31:55] It's like, if you need to comply with a piece of legislation, then take a snapshot at a point in time at whatever frequency is required and, you know, submit those results or whatever the law requires.

[00:32:06] But when it comes to, you know, mitigating risk and understanding all these other, you know, implications way beyond the law, you do need, you know, sort of continuous, you know, monitoring and observability to understand exactly what's happening.

[00:32:21] And make sure you don't have any weak links or whatever, similar to, you know, cybersecurity or privacy, I would imagine.

[00:32:28] Let's get back to manufacturing and automation, right?

[00:32:31] And imagine the AI agent I'm creating is about designing or putting together, assembling the frame of a car.

[00:32:41] One company will decide to just like put a piece of aluminum, whatever they use, the metal they use, and to solder it.

[00:32:47] The other one will actually decide to like glue them together, those special glue, blah, blah, blah.

[00:32:52] Okay.

[00:32:53] Within the framework of the law, it might be possible.

[00:32:55] Now, what you cannot do is just do like duct tape.

[00:32:57] Okay.

[00:32:58] That, the long term you know.

[00:32:59] That's kind of the same thing.

[00:33:01] But now it's just how maybe gluing it is much faster, but will create like over time some recall of all the car you make because some of the frame have been collapsing.

[00:33:13] That you cannot know in advance, but you have to monitor that to be able to maybe change it.

[00:33:18] And I think tomorrow's organization will be the one who are creating.

[00:33:22] It's not, it's not who are using more or less agents, but who are defining what the agents are doing, want to get one with the other or like on themselves the best.

[00:33:32] And it will kind of like, I think encapsulate somehow the organization way of doing things.

[00:33:40] The same way one organization is not the same culture than another.

[00:33:43] An organization is not competing in the same way as another or has a different approach to markets or a different view of the world, different like strategy on how they bring product or service to market.

[00:33:55] It's exactly the same.

[00:33:56] So keeping an eye on it.

[00:33:58] Yes.

[00:33:58] And actually there's a very, very, very strong business guess around that is my computer is better than I am because of that type, that type of AI agent.

[00:34:09] Okay.

[00:34:09] That is clearly like we all agree that's the best we can use for that layer of whatever operations or process that we have.

[00:34:18] Then let's keep an eye on.

[00:34:20] It's comply with like the legal framework, blah, blah, blah, but also that the better job.

[00:34:25] And that will just go like the competitiveness will just move to another layer.

[00:34:29] So now like the different assembly of agent AI.

[00:34:34] And that's really what it is.

[00:34:37] And that actually I think will define really what our organization is better than another.

[00:34:44] Our HR function is better than another.

[00:34:47] Our tool, talent, intelligence, administration of data is better than another.

[00:34:52] It's not just about the agent AI.

[00:34:54] It's how you put them together.

[00:34:55] And what sort of, yeah, filter of, I would say like personality do you give them?

[00:35:02] Yeah.

[00:35:02] That gets into a whole nother area of how these things are sort of programmed.

[00:35:06] Because I happen to be in a sort of AI agent kind of boot camp right now.

[00:35:12] And we're learning how to, you know, make sure it's speaking in your sort of brand voice, right?

[00:35:18] If it's communicating with clients and prospects and customer service or whatever.

[00:35:23] Like if you're going to apply an agent, then don't make it too, you know, sort of robotic, too generic.

[00:35:31] Don't make me choose from one of these five, you know, topics and only follow this set of rules.

[00:35:38] I mean, that's like a old school conversational chatbot at this point.

[00:35:42] I think a lot of people already know how to, I mean, I have probably a good opinion on how to do that well inside, you know, bigger organization.

[00:35:49] I'm thinking like, say, people in marketing.

[00:35:52] Look, when I'm...

[00:35:54] Hey, everybody.

[00:35:55] It's Libby again with fearlessness.

[00:35:57] So what's fearlessness?

[00:35:59] It's that underlying grit that empowers us to forge ahead, even when hope seems distant.

[00:36:03] It's the courage to walk through those fires of hell, knowing that we're going to come out better and stronger on the other side.

[00:36:10] Stay tuned and learn how to get fearlessness.

[00:36:13] You're still using slides, right?

[00:36:15] And she worked at IBM when you open PowerPoint, you know, your PowerPoint will be designed in an IBM way.

[00:36:21] And if I work for computer and for big name in a different way, that's, you know, that transcribe and carry your company, I don't know, logo, like, you know, frame, like, in a certain way, like, visual image, whatever you want to call it, right?

[00:36:36] This is exactly the same.

[00:36:38] Just to just take it instead of, like, instead of the platform being, like, a PowerPoint, now becomes, like, you know, a way to interact with, like, a client.

[00:36:46] It can be a chatbot.

[00:36:47] It can be whatever you want.

[00:36:48] But it's exactly the same principle.

[00:36:49] The IBM one's an interesting example because the brand guidelines about how you can use logos and colors and, you know, whether or not you can put another logo next to the IBM logo.

[00:37:01] I mean, it's incredible.

[00:37:03] The whole team's putting together just the rules about how that needs to work.

[00:37:09] And it was, when Watson came out, it was probably worse because they just didn't want to, it was a risk mitigation strategy in part because they didn't want to dilute the branding of that.

[00:37:23] But I think we understand each other.

[00:37:25] It's just like the same kind of question will arise.

[00:37:28] It's just like applied to AI or something that's with AI.

[00:37:31] So, if you control that, how do you make it, like you said, not to robotics, how do you make it, like, so we carry, well, who you are as an organization or what you are as an organization.

[00:37:41] Yeah.

[00:37:42] If I were to sort of summarize some of the key, you know, thoughts we've been talking about, it's like, forget about what the AI is capable of for a moment.

[00:37:53] And just think about some of the fundamentals that go into making these things successful, right?

[00:38:01] Like, I've always contended that, you know, your data and analytics maturity in itself is a huge, you know, benefit and competitive advantage when you start to build with AI or, you know, just pull your data and some of your proprietary information into an AI solution that you may have procured from a solution provider.

[00:38:22] But, I mean, that's the foundation for anything that any output you're going to get from that AI.

[00:38:30] If you don't trust the data and some of the core components of that, then all bets are off.

[00:38:36] And that is especially true for, you know, decision making when it comes to, you know, AI solutions in the talent space, right?

[00:38:44] So, certainly looking at, you know, DEI, I mean, we can debate, you know, whether DEI is properly, you know, supported.

[00:38:54] We can, you know, talk about how corporations are reacting these days to, you know, DEI initiatives.

[00:38:59] But the fundamental principle of, you know, fairness and impartiality and bias mitigation all goes back to the data.

[00:39:10] And so, you've got to look at that before you start blaming the AI.

[00:39:14] How would you do that or not?

[00:39:16] And, you know, by the way, like on the DEI and like the idea of like, you know, responsible AI.

[00:39:22] One thing that came to mind before we started recording on that is people tend to think about responsible AI as, you know, how do we use AI responsibly?

[00:39:32] But I think there's another way also of looking at that is what is our responsibility in actually deploying AI and the topic of DEI, or in the topic of just simply having, you know, and I'm just like the use case for the hiring or maybe internal mobility.

[00:39:50] Is how AI can actually help an organization to just not do the job of the recruiter slightly faster, but really like get to every candidate and try to figure out like deeply how everyone is not as a fit, but will work all the time and what they can bring and better.

[00:40:09] Not just like giving a scoring on how the candidate will be a good or not, but it's kind of like giving more chance than what a human will be.

[00:40:18] You know, nowadays, like if you look at a lot of job posting, you know, talent intelligence or workforce planning, well, you have to have like so many years of talent intelligence.

[00:40:28] I mean, you have to come from there.

[00:40:31] So people like you and I might not have the same chances and still be working there and are pretty successful.

[00:40:37] So how come?

[00:40:38] Well, if an AI could actually ask me more questions that they read that about your resume or whatever your things, your online profile.

[00:40:46] Like tell me more about it.

[00:40:48] And I can from there, like in asking me more questions, just be, I think the patient that a human might not have as candidates, I will probably really appreciate that.

[00:40:57] I will kind of like give myself a chance to defend myself without so much pressure.

[00:41:02] And actually talking to a system that I will actually like try to really make a balance between like, you know, who I am, where I'm coming from and not, you know, and kind of get out of the usual standard, you know, of like, that's what we're looking for.

[00:41:17] No, that's not, you're not looking for a candidate that has been doing that, but that you're looking for a candidate that could do that another time.

[00:41:25] And when that will change, it will be able to go on with the change.

[00:41:28] So do you want someone who actually has been doing exactly the same for 20 years or someone who actually has been moving a lot for 20 years?

[00:41:34] If the role is potentially, will potentially be exposed to a lot of changes, something very rural, you want someone who's been actually exploiting itself, something very rural.

[00:41:44] And it's, you know, there's not one good solution to that.

[00:41:48] But I think, you know, back to the topic, I think we all have a responsibility to use AI in those moments and to use AI well, to use AI to its full potential.

[00:42:01] It's sure, it's sure, it's about like, you know, do we have the right data?

[00:42:05] Do we build the right algorithm so we can put people out?

[00:42:08] I said, yes.

[00:42:09] And what about, we will also make sure that we put everyone in.

[00:42:12] And that's the responsibility, I think, we'll have when it comes to AI.

[00:42:17] It's not about like shying away from it and saying like, well, before we use that, let's see what, you know, before it created problems.

[00:42:24] Like you can, I mean, you know, developers have been using Sandbox for a while.

[00:42:28] But just take a sample of your data.

[00:42:30] Do that on the site.

[00:42:31] Test it.

[00:42:32] Try it, you know.

[00:42:34] It's fine.

[00:42:34] It's not public.

[00:42:35] It's there.

[00:42:36] It's not about hiring someone.

[00:42:37] It's just about to see like, does it work?

[00:42:39] Can we be better?

[00:42:40] Can we hire?

[00:42:40] And it's not about hiring the right, the best people.

[00:42:43] It's about hiring the right people.

[00:42:45] I think, you know, it charges a lot into hiring or using the right people.

[00:42:49] It charges a lot into like, we want the best of people.

[00:42:52] No, we don't want the best, we want the right.

[00:42:54] Yeah.

[00:42:55] No, there's so many good nuggets in all of that.

[00:43:00] I think you called out a couple of things.

[00:43:03] First of all, I think responsibility, you've got to be responsible by design, right?

[00:43:07] And that starts with whoever's, you know, building and developing this.

[00:43:13] It's not just the solution providers anymore because like we were talking about, anyone can now build a GPT or an agent.

[00:43:20] And so now everyone sort of takes on some of that responsibility.

[00:43:25] It would seem easier to save than to do it.

[00:43:28] But yeah.

[00:43:29] Yeah.

[00:43:30] But the other is, you know, just even going back to your point about, you know, the role as it's defined.

[00:43:35] You know, we spend so much time worrying about matching, you know, job descriptions to CVs.

[00:43:42] And there are so many problems.

[00:43:44] I mean, we could spend a whole nother hour talking about why that's such a bad idea.

[00:43:48] Not just because, you know, now it's, you know, an AI generated JD versus an AI generated resume.

[00:43:53] But the fact that the role, that role is going to invariably change in a pretty short amount of time.

[00:44:02] And you need to understand whether that person has the ability to learn and grow and adapt as the role changes.

[00:44:11] And that role could change for a variety of reasons.

[00:44:14] First of all, that one bullet point that says, you know, other duties as assigned could easily blow up and take up a significant, you know, double digit percentage of the work.

[00:44:25] Or you see all these changes in your business model, in the markets that you go after, in the way that you do your manufacturing and all of these other elements that go into the sort of variability of the role, which goes back to your very first point.

[00:44:44] Around the sort of agility and adaptability of, at every level, right?

[00:44:50] From an individual all the way to the organizational level.

[00:44:53] Yes.

[00:44:54] We could spend another hour talking about matching things and stuff.

[00:44:58] But at the end of the day, you know, there's always like, if you think that it's a, just when an organization is trying to find the one person for that role.

[00:45:06] Whereas like a person is trying to find the right best job, you know, and the chances that you find really like the exact perfect magic.

[00:45:14] Match is very small.

[00:45:17] But now we start hiring a lot of people here and there and people moving and stuff like that.

[00:45:21] So maybe it's just like, you know, different ways to look at it.

[00:45:26] But yes, yes, yes.

[00:45:27] And it's not just, you know, I mean, to be like hiring or like internal mobility.

[00:45:33] It's one of the many, many, many, many, many, many use cases there are.

[00:45:37] And actually, there's a lot of use cases that people are not thinking about where I can be useful.

[00:45:41] Use cases that do not exist necessarily today that actually mix up business or other function priorities with HR priorities.

[00:45:52] Before we were mentioning a strategic workforce planning.

[00:45:54] When you do a strategic workforce planning, a lot of business input is necessary, a lot of like market knowledge is necessary.

[00:46:04] It's data that lives somewhere in the organization.

[00:46:07] They might, let's assume they can be somehow available.

[00:46:11] AI can actually help HR or anyone else to actually just like, not say like, well, strategic workforce planning should belong to HR.

[00:46:20] No, it doesn't belong to HR more than that belongs to business, more than belongs to finance or anything like that.

[00:46:26] It belongs to the organization.

[00:46:27] So how to balance that goes information to create like, maybe it needs to be called something different.

[00:46:34] But to do something that will actually be meaningful, purposeful and usable.

[00:46:41] AI can probably help with that.

[00:46:43] Like, hey, agent AI.

[00:46:45] Create something where, you know, balance, you know, skill view of the organization to the financial view of the organization to the market vision view of the organization to create a workforce plan.

[00:46:56] That makes sense.

[00:46:57] Over the next 18 months.

[00:46:58] And in that next 18 months, we want three scenarios.

[00:47:01] One especially, what was the best in the market.

[00:47:04] And that's one, what was the worst in the market.

[00:47:05] Like, AI can do that.

[00:47:07] We can.

[00:47:08] Or could do that.

[00:47:09] But if you just look at it like, just do something HR and not look at the rest of the world.

[00:47:15] It's good, but it's not great.

[00:47:18] And is it so useful?

[00:47:19] I'm not sure.

[00:47:20] Well, I do think that not enough organizations are doing strategic workforce planning to begin with.

[00:47:27] But, again, that's probably a whole other conversation.

[00:47:32] But you're right.

[00:47:33] You know, AI can help with that exercise.

[00:47:37] AI is a factor in how you execute it, right?

[00:47:42] You need to understand what AI will continue and what capabilities it will gain that you need to incorporate into, you know, everything from the work you can accomplish to, you know, the skills gaps and things like that.

[00:47:57] But, yeah, I still think, as we talked about, a lot of it just comes down to, you know, behavior and leadership and culture saying, we need to do this for our own competitive advantage, for our own survival.

[00:48:13] These are things that we need to do because we can't be on our heels.

[00:48:17] You know, we want to be a thriving organization in all sense of the word, not just because our financials are better or like for growing or this and that.

[00:48:27] But because our people are so great in that organization, because people want to work with us, because other organizations want to partner with us, because people are looking at us as like, wow, this is where innovation is coming from.

[00:48:38] There's so many reasons, you know, and they're not just like pure financial or business oriented.

[00:48:44] But at the end of the day, it's like, we want to thrive.

[00:48:47] We want to start an organization.

[00:48:48] We want people to be proud to work.

[00:48:50] We want people to be proud to work.

[00:48:52] We want people to be proud to work.

[00:48:52] And guess what?

[00:48:53] They might at least stay longer.

[00:48:55] Wouldn't that be something?

[00:48:55] And Remy, I'm going to leave it there because otherwise we're going to talk for another hour.

[00:49:00] So I want to be respectful of your time.

[00:49:02] This has been great as always.

[00:49:05] It's great to catch up.

[00:49:06] And thank you for sharing all your insights with my audience.

[00:49:11] Really appreciate it.

[00:49:12] Well, it was my pleasure.

[00:49:13] Thanks again.

[00:49:14] And we appreciate you giving me a voice here.

[00:49:16] Of course.

[00:49:17] Of course.

[00:49:17] Anytime.

[00:49:18] All right.

[00:49:19] Thanks again, Remy.

[00:49:19] And thanks everyone for listening.

[00:49:21] We'll see you next time.