Speaker 1: 0:03

Hi and welcome to the Data Topics Podcast. My name is Ben Team Lead Data Strategy at DataRoots, and in this upcoming series of episodes, I will be talking to a range of AI leaders coming from different backgrounds. In this episode, I'm joined by Jacques Janssen. We will talk about how the role of AI has changed throughout his career, about leadership lessons and styles, and also about rediscovering humanity in the era of AI, which is also the premise of his new book. I actually asked Chet GPT to come up with an introduction for our new guest. Today we're joined by Jackie Janssen, belgian author and chief data officer at CM. Known for his practical approach to digital transformation and AI, he's the author of AI the Hype for Bear, a book that strips the bus from artificial intelligence and makes it accessible for real world impact. With a career spanning leadership, entrepreneurship and thought leadership, jackie brings a human first mindset to technology, and we're excited to learn from his journey. What do you think about that, shaki?

Speaker 2: 1:07

Well, it's quite an introduction. Thanks for inviting me and also for the introduction as well, ben.

Speaker 1: 1:13

Is it still correct, because it says Chief Data Officer at CM.

Speaker 2: 1:17

No, not at the moment. I had to quit my job at CM because I could not combine it with the things that I'm doing at the moment as a consultant. So I certainly think CM needs a full-time CDO and I was not able to do it anymore.

Speaker 1: 1:37

Okay, something that we'll try to tackle later on in this episode because I'm very curious, great tackle later on in this episode because I'm very curious, great, um, maybe start at the beginning of your career, because, if I'm not mistaken, you studied it and then started your career at kbc yes, at that time it was called credit bank.

Speaker 2: 1:56

But yes, I started there, I think, I don't know, 1996 or so.

Speaker 2: 2:00

So you make me feel like a dinosaur when I was born oh great, you make me feel like a dinosaur, but at that time, uh, yeah it. You cannot compare it to the it of now. It was certainly a support function. Uh, even, you had the developers sitting in one room and then you had the analysts sitting in another room who are doing things with requirements, and then a full analysis came to the developers and you had to start what we call the technical design, which was done by technical designers comparable to a lead developer at the moment. And then we started coding. It was, I don't know.

Speaker 1: 2:41

I think I started with COBOL and it was PL1 as well, on on ZOS, on IBM mainframe so a lot of things change a lot yeah, yeah, obviously, I think after that you took on roles as coach and consultant, and then you returned to KBC 10 years later yeah, that's right yeah what has changed them and what didn't change? Was there a big change?

Speaker 2: 3:08

well, yeah, certainly there was a big change. But, yeah, you have to know, I started there but I stayed also there for a long while I think 11 years, I don't know, 10, 11 years. Then I went to, uh, indeed, some consultancy and, returning back, the thing that, uh, that certainly changed the most is the composition of the team. Set was more a multidisciplinary first step in agile working. When I came back at the end, it was certainly a full scaled agile thing. And also, the black screens with the, the green letters at the terminals, they disappeared. It was more, yeah, well, really full experience that you need. Also, technology changed a lot. It was Java stack and for me and my team it was really an inspirational environment as well. We had a lot of possibilities to experiment, to innovate on a Python ecosystem, on AWS. We created our own environment to do data science. So, yeah, things change a lot.

Speaker 1: 4:18

So the technologies changed, the organizational structure changed. How did that enable the cooperation with business?

Speaker 2: 4:27

well.

Speaker 2: 4:27

Well, I did this.

Speaker 2: 4:29

One of the major things uh, you wanted to talk about leadership, the the major thing that you have to change is certainly not technology, it's uh, it's the people, it's the culture that you create and, in evolving to a more it business collaboration, what was certainly, what certainly was the goal?

Speaker 2: 4:52

Yeah, you have to create more culture to enable that and make it a win-win. So, when I started, with the analysts in one room and the technical designers in one room and the developers in one room, you really need to tear down the walls to have this kind of a collaboration, and that happened. So that was certainly a huge change, also on the level of vision, to be honest, to, to my opinion, a bank is, at the moment, an it company. It's using it to create the services. When you go back 20 years ago, it was certainly the other way around. It was not at all an enabler, it was just technology that needed to be there so they really want to invest in it being a competitive advantage it is a competitive advantage and when you add data and ai on top of it, it certainly is a huge advantage.

Speaker 1: 5:55

Yes, I agree you talked about it being a silo at first, being really a support role. You also talked about culture. I think many companies want to have a more data driven culture, but they find it very difficult to launch initiatives. Yes, you can have a training or you can talk about ai, but can you think of initiatives that really helped shaping that culture at kbc back at the time?

Speaker 2: 6:20

well, I, I really think you, you need to do the bottom-up thing, what people are talking the most about, but strong leadership, a top-down setting, a compelling vision towards the people, creating teams who work towards that vision with strategy, with concrete projects, in a pragmatic way that creates kind of the culture. Also, every time again, say it needs to be data driven. Why are we not doing it that way? I creating really a dna around it that makes it that that's certainly one of the things that makes culture work Creating champions, success stories on the stage. These are the things that you need to set, because then you are, from a leadership perspective, expressing what you expect from people.

Speaker 1: 7:19

Nice, very nice initiatives, very inspirational. You progressed your career a lot at kbc. Could you talk a bit about that for the younger audience that listening? Um, how did you progress your career and what was the main? With the main, yeah, what was the main reason for you progressing your career? Is it because you were more oriented towards business? Was that a crucial aspect, or was it something else?

Speaker 2: 7:48

what? What certainly helped is that I could explain complex concepts in a in an easy way. That certainly helped a bit of luck as well. I started and two years after I had some ideas which I expressed and which were picked up by the leadership so that I could develop my own ideas, creating innovative projects as well. So that helped see the chance and take it, and that's only one of the things that that you need to do be very collaborative with the people that you meet on business level, on it level, even on on the infrastructure level. Create kind of a team where you yeah, you are very collaborative. Um, even at that time, uh, my, how do you say it? My, my thinking, my, my thinking was not that waterfall as a bank was, at that time I really went for fast deliveries. Uh, every time. What's the return, what's the value? Creating these kind of things helped me in in my career.

Speaker 1: 9:11

So kind of a business mindset, an entrepreneurial mindset, in what was certainly not an entrepreneurial environment and then, when you became a leader, what was your leadership style and how did it influence the business?

Speaker 2: 9:27

it alignment within the company oh, leadership styles we can talk hours, but leaders, I think mine is more supportive. I really help assist people. If they do not find a solution and I cannot find it, I will find a way with the people around it. So really, really helping and supporting.

Speaker 1: 9:52

So really focused on collaboration and a certain degree of autonomy too.

Speaker 2: 9:56

Autonomy is maybe the key word in the things that I said, because I even experimented with completely self-autonomous themes.

Speaker 1: 10:07

How did that work out?

Speaker 2: 10:08

Well, it worked out well, but you learn a lot from it as well, because some people need kind of a structure, need some guidance, but others don't, and it's not a black and white story. But yeah, it certainly helps if you give people autonomy, they feel involved and they come up with the right and use new ideas, and and you also need to create a kind of an environment where people can fail if they they need to these kind of things, it helps I 100% agree with that.

Speaker 1: 10:44

Um, afterwards you became a cdo at cm, which is a very different industry, I think maybe in some way it isn't. But, um, how do you compare working for different industries and what are like the similarities and the differences, uh, in working for those different companies and industries?

Speaker 2: 11:02

yeah, what certainly was different is the regulatory environment where a bank is working in or where a mutuality like cm is working in. It might sound weird, but the stakeholders of mutuality it's more complex it's government, it's RISF, it are the practitioners, the medicines. It's quite a complex environment to deal with, so that was certainly different. What was very good is that we also there and that's kind of a similarity with my last years at KBC we also there created kind of a similarity with my last years at KBC. We also there created kind of a compelling data, ai driven vision and brought that to the people try to well, try to guide them to get there. So that was quite the same thing. What was also different for me at CM but I also had that at some startups that I did as I had to create, create a team from scratch, uh, which was really a well, a nice thing to do, because then you choose the people that you are working with.

Speaker 1: 12:17

What are the first position that you want to fill.

Speaker 2: 12:20

What was the first position?

Speaker 1: 12:22

Yes, in a team. I can imagine you have a range of skills more data science oriented, more translator oriented.

Speaker 2: 12:28

It is documented in my book. Even it was a data scientist, someone who could really create innovative things based on data, and it was the first thing I filled in. I started in May, in August. The first thing I filled in. I started in may, in august. The first data scientist started in september, the second and then, I think, after three years, we were with about 40 people and that's because you want to have a pock first to show value, to show value leadership really showing value.

Speaker 2: 13:00

Uh, and and not on the basic things, because I one of the other things we started immediately is incorporating the BI and the data team, Because you have to have the data and people who are doing these kinds of jobs. Creating data pipelines or even the ETL processes to get data available to users is one of the things that is well. It's underestimated, I think, in companies.

Speaker 1: 13:30

And then in terms of the operational model, the culture, but also the typical AI use cases within the industries. How does that compare to each other? Banking versus mutualities.

Speaker 2: 13:43

Well, there we had a huge difference in what was already there and what you had to create. Normally, when I come in a company where there is also I some because now I do a lot of consultancy and then you come at a huge company, small company, some have data and ai divisions, some don't, and normally, after some kind of assessments, almost seven out of ten in in in belgium we start with more first technology focused team, teams and the. The reason for that is because you need to to have kind of standardization, some kind of productive industrialized flow, because once that is there then you can go very fast to business. So what I have then is I kind of transition from the technology point of view to a more business oriented view. So that's kind of a shift in teams and at CM.

Speaker 2: 14:47

That was certainly one of the things we did, but also in most of the startups, and even without AI, if it's about getting people working in an industrialized way, almost as a factory in IT, industrialized way, almost as a factory in IT, so you need to start from this technology stack and then you can evolve towards a business stack, because otherwise architecture, infrastructure, the software that you are using, is getting kind of scattered all around the company and then you don't have standardization and costs will increase. So you try to get that as low as possible it's a bit related to mlops, I assume yes, well, mlops is one of the ways to fill in that.

Speaker 2: 15:35

You can do mlops once the teams are there and your mlops procedures are installed. It's very difficult, for instance, if you would have seven different IT teams that are focused on one on HR, one on finance. So, based on the business divisions that you have in a company, and you would then start installing MLOps procedures, we'll have a lot of discussions, a lot of flavors, and it's easier to have it in place at one team and then install it at the other ones. Or, if you have the possibility to start from a technical team, then you install it in the rest of the company.

Speaker 1: 16:23

You just talked about factories. It brings me to the product thinking what do you think about that?

Speaker 2: 16:31

I think that's a very good way of working. It needs some kind of maturity as well. So, also there, you you need kind of a fundament before you can go to this factory way of working, to this product thinking. But, to be honest, from my point of view, I think it's the best way to go to. That's what you also see in agile teams skilled agile teams. You need to introduce product thinking, and that goes very broad. That can be data products as well. And that goes very broad. That can be data products as well. So, yeah, I like product thinking and design thinking also around products.

Speaker 1: 17:12

Yeah, indeed, because you want to have the highest possible level of quality for your AI product or data product, which brings us also to governance. I heard you talking a lot about AI councils. I think you mentioned that it's very important to have one. What do you consider a very well working ai council?

Speaker 2: 17:31

yeah, I, I think it's needed, certainly for for bigger enterprises to have it. But yeah, the idea of an ai council even if you're a small company, you would like to do it and you need to combine, indeed, some skills there, and I'll start with ethical skills. I always, in Dutch, I start my talks with technology, whether it's digitalization or ai. Is this something that is in line with your company values? And the ethical guy is the one who needs to guide that. And then you have compliance and legal, which needs to be there. So if you have that installed, I prefer the CEO of the company to be in the council, because that's the top-down guidance that you need. That's the support that you need. That's the one who can make decisions on all levels.

Speaker 2: 18:39

And then the business lines need to be there, but I think with six, seven people, it include in it, because if you want to industrialize, you need them. That's nothing. That's not a thing that you can do from an ai competent center of excellence, and even you don't want it because they they need to be busy with ai, with AI technologies, ai solutions, getting the business challenges to the department and find solutions for it. Don't let them bother about how these things integrate in an IT environment, so there needs to be a strong collaboration and that's why the idea of product thinking is a good idea, because then, every time you need you, you are sure that you want to go with your things to production, which seems to be a struggle for a lot of companies. You create a multidisciplinary team with your data scientist, based on guidance from the ai council, with your it, with someone of legal doing the follow-up and with your business people, and if it's multidisciplinary, you will succeed.

Speaker 1: 19:53

To make it very concrete, the AI council. What is the frequency of meetings and what are typical topics on the agenda? Because, for example, you want to also include these people for use case discovery, but is it something you would do during the AI council meetings or is it something that's separate from that?

Speaker 2: 20:09

Yeah, it's separate. If I take a look at the AI Councils which I guided, I think it's almost seven, eight years. I'm getting old If I take a look at that. What I always did is I start with an inspiring case outside of inside the company I don't mind but that they are aware of the possibilities that are there or where we need a decision based on the advice given by ethics, by IT, by compliance, and then you have this AI council deciding on these topics. Lastly, topics like your AI policy, the hurdles that you have in a company they need to be addressed there as well and AI culture for sure.

Speaker 1: 21:11

So it's really about making the things the AI team is doing visible.

Speaker 2: 21:15

Tangible and visible.

Speaker 1: 21:16

Tangible and visible, indeed, and then also steer it in such a way that it's ethically and regulatory compliant.

Speaker 2: 21:24

Well, at least it is compliant with your own company values okay cool.

Speaker 1: 21:31

It's a bit intertwined with the lessons you share in your book, ai, the hype for bear. Of course we need to talk about it. It's gained a lot of attention in the past year. Your life changed. Yeah, I can imagine yeah, my life changed.

Speaker 2: 21:45

Yeah, uh, because, yeah, I was chief data officer at cm when I wrote the book and after writing it, I was asked a lot to do keynotes to the lectures, to do board advice, to do some basic consultancy as well and and well, as I said at the start, I could not combine it with my job as a CDO, so I had to resign, which was a bit pity, because I really liked working in the health environment. We were doing very nice projects. But, yeah, now it's completely different. I'm doing these keynotes but also advising companies, so it's a broader perspective and I'm selling books.

Speaker 1: 22:34

I'm sure they miss you at CM. So now, in your new role, did you have like inspiring new lessons? Or because I can imagine you talk to different people, different industries, different companies is there something that's similar between all of those different companies and people?

Speaker 2: 22:52

um, well, well, what's certainly similar is a lot of the companies are still searching. They see the potential of ai most of the times on an individual level, so on personal productivity, they want to scale that to teams, to companies, and that kind of seems difficult to do. But I compare it to previous waves we had, when you take a look at the internet wave also there it were first some individuals who could find their way to find things on the internet, to find solutions on the internet. Same happens with the mobile phone and now you see it with AI. It's kind of a struggle.

Speaker 2: 23:39

The second one is what we already talked about a bit. If they have solutions or challenges they can solve with AI, it's kind of hard to get it into production. So that's one of the things that I see. I recommend to start with processes. Take a look at processes, preferably your core processes, analyze them and look there for AI possibilities, gen AI possibilities, and then you will see because it's your core process that you will gain anyhow something, and then again potentially work with AI champions or even yeah, that's to set a culture, that's to set a culture.

Speaker 1: 24:26

Why did you write your book in the first place? What was a key moment for you? When did you decide I really need to write a book?

Speaker 2: 24:31

yeah, well, I, there were a few uh things that the first one was certainly when I was at a conference and that I heard someone talking about the eye, kind of as the magic wand that solves everything, and then I said, okay, but it's really not like that With 10 years experience, at that point in time I really had to do something with that message because that was not reality. Secondly, it was half a year after ChatGPT was there when I wrote the first pages, and that's because I saw people doing things with ChatGPT. What I questioned why are you doing that? Why are you looking for the birthdate of Napoleon with ChatGPT? You can find it on Wikipedia.

Speaker 2: 25:24

All this generative it's. It's certainly a good solution for a lot of problems, but it's also not the right solution for a lot of challenges that we have. And at that moment in time, it felt to me that everybody uh, going on this hype and going to chat, gpt and generative vi to find solutions for everything, which certainly was not the case To me. It's still well, it's statistics on steroids, this ShedGPT or other GenerativeVI tools, so a lot of things going on things.

Speaker 1: 26:09

Uh, well, it felt to me okay, uh, let's write down how you really can do it, uh, and why it works.

Speaker 2: 26:14

If you do it the right way.

Speaker 1: 26:15

Now people are talking about agents, so even I think um, not everything that's mentioned about agents is true, and there's a lot of hype, and I think it's in a way, similar to what has happened to generative vi in the chadgety moment in, I think, 2021 or 2022 how do?

Speaker 2: 26:31

you look at that yeah, uh, well, you can compare it to what you are saying. Uh, it's a bit the same. Like we do with almost every technology, we overestimate it at the short term, but we underestimate it on the long term. To be quite honest, I'm doing a lot of experiments at the moment with AI agents, and every day again even last weekend, I started experimenting with a development agent and every day day again, I am surprised of the possibilities and then I'm thinking, okay, but it's only now. What will it be after two years?

Speaker 2: 27:15

So I think we overestimate it and and that's due to we, we seem when we see a new technology that's going, uh, that's doing some magic, it seems that we are always, uh, forgetting that there are human beings who need to use the technology and there are legal things that you need to find solutions for, and seems that we forget those two and they are the slowest parts in change. It's not the technology. Technology is going fast. So, the possibilities of AI agents and whether it will be a really breakthrough for generative-based, ai-based technology yeah, I think so. It's a game changer.

Speaker 1: 28:03

It makes me think about what you said earlier, that you want your teams to be very autonomous. You could argue that I mean companies are looking into having teams of agents that would work autonomously too. How do you?

Speaker 2: 28:15

look at that. Well, we will first. Well, I have set up a farm of AI agents and it can work very well, but we will have to go through a first phase where we need to learn people in teams working with team members who are AI agents, and that will change also the way that agile teams work, because how would it work if one of your team members, which is an AI agent, doesn't need a sprint of 10 days to solve an issue? He will do it in half an hour, and so you need to think about how will this collaboration work. And also there, it's not about the technology, it's not about the ai agents, it's about the people working with it and what will happen on the legal part of it.

Speaker 2: 29:09

So the adoption will not be. It will not be a technology adoption and certainly on the human part, we have to solve a lot of things. As I told you, I'm working a lot with AI agents. I'm starting up maybe between 10 and 15 every morning to do some tasks, but I cannot handle the output and it even confuses me by moments on. Okay, now I have the output and what to do first.

Speaker 2: 29:41

It's outspeeding us? Yeah, indeed.

Speaker 1: 29:44

So we have to find solutions and we, we have to find solutions and so we have to find a way in how to collaborate with these agents yes yeah, um, so you're staying ahead of the curve by already doing some thinking about that. On that, um, is it true that you're writing a second book? Yes and what is your second book about?

Speaker 2: 30:01

it. It's well, in Dutch it is AI, de herondekking van de mens. So in English, rediscovering Humanity, and also there. Both books will be in English from October on.

Speaker 2: 30:19

So Rediscovering Humanity and where I start from is well, we kind of identified ourselves as human beings with the knowledge that we have, even from the start. What school, kindergarten, what school you are going to? It's all about knowledge and getting knowledge to you. But what will happen if knowledge is not important anymore Because the AI is addressing all the knowledge and knowledge questions? If that is the case, what remains of us? We are still curious, we are still empathic.

Speaker 2: 31:00

So there are a lot of things that we are still having as human beings, which is completely different than the things that you do with AI. I identified nine things where an AI well, he can try to fake it, but he will never have it like curiosity, for instance. And then the second part of the book are all things where ai is already ahead of us. Let's not discuss about it anymore because it's waste of time. We have to adopt, we have to take time for the adoption, but it will be there. And if you combine both human strengths, human characteristics, with the strength of ai, what will coexistence between ai models, ai agents and human beings be?

Speaker 1: 31:56

that's super relevant, given the agentic ai and generative applications we talked about, but also ai in general. What are the key messages of the book or who should read the book?

Speaker 2: 32:07

everyone should read the book.

Speaker 2: 32:08

Yeah, of course, especially now it's in english we have a lot of english listeners so, yeah, uh, well, I key messages are certainly, uh, that for sure, even if ai agents are there and doing a lot of the work, it, it will not mean that we go to a dystopic world. It will also not mean that you can live in the utopia, to be in your hangmat and on the beach and only talk with friends. You still have to work. It will be a balanced world where we will live together with these models and, to be honest, we are already doing that with these models and, to be honest, we are already doing that.

Speaker 2: 32:49

You drive, probably with ways, which is living together with ai and and other things. People who wake up with an alarm that is flexible and based on ai algorithms, saying, when you have to wake up, these kind of things, they are already there and it will be more, more, even on your work, you will see a lot of these things. So I, that's certainly one of the key messages, but also there I go beyond the hype take a look at how it can can mean an advantage for you and to to be busy with the things you really like, even on your work.

Speaker 1: 33:25

You often mention that if you would ever gain in personal productivity thanks to these agents or other AI tools, it's not just to have an extra coffee at work. But on the other hand, do you think on a societal level, it will help people in getting more time for meaningful things like family sports and so on? And how will that?

Speaker 2: 33:51

I'm sure, but that has to be a combination of what we want as a human being, and I meet a lot of people who are looking for ways to get out of the pressure. People who are looking for ways to get out of the pressure. We see a lot of burnouts, a lot of things happening society and I think ai, if it is used in a right way, if we have the right framework to do it, if we don't accept the pressure of doing even more now with ei, then I think it really can be a solution interesting.

Speaker 1: 34:30

When will the book be published?

Speaker 2: 34:32

uh, september, september something to look forward to um.

Speaker 1: 34:36

Do you also discuss, for example, the impact that it will have on education?

Speaker 2: 34:42

yes, education, health, uh well, hr as well, uh, but on education, yes, because I think we we need to take a look at what we are doing at schools. Uh, I advise a lot at schools, uh, and sometimes you see them fighting against this ai and sometimes you see them embracing, working with ai. But if you then take a look at the nine things that we we still have as a human being which we we can strengthen critical thinking is one of them and we from from well, from kindergarten, we need to help these kids with critical thinking. Avi is there. Same about curiosity or creativity. There are a lot of things where we still are in front, a lot in front, and these are skills that we really need to help and support and engage for in schools.

Speaker 1: 35:50

I will definitely buy a book. I'm very curious to the second part to understand where AI is even better than I am today already, but also very much interested into the human aspect of AI and what we are up to as a human in the coming years. I think we can wrap up the session. I want to thank you very much for being here. It was very interesting and I want to wish you the best of luck with your existing book and the new book and everything you're doing in supporting companies and humans in finding themselves and finding the right way to approach this new way of living and these new technologies.

Speaker 2: 36:29

It was my pleasure and thanks for inviting me.

Speaker 1: 36:31

Thank you very much.

Speaker 2: 36:35

You have taste In a way that's meaningful to software people. Hello, I'm Bill Gates. I would recommend TypeScript. Yeah, it writes a lot of code for me and usually it's slightly wrong.