Lead With AI
15
min read

Beyond the AI Pilot: How PwC Scaled Adoption Across 6,000 Employees

PwC Netherlands' Marlene de Koning reveals strategies for scaling AI adoption from pilot to enterprise with workflow integration and AI influencers.
Published:
August 13, 2024
Last updated:
October 15, 2024

Also available on:

Future Work - Listen on Spotify
Future Work - Listen on Apple Podcasts
Future Work - Watch on Youtube
beyond-ai-pilot-pwc-scaled-adoption-6000-employees

🎧 Listen Now:

In today's episode, we speak to Marlene de Koning, Director at PwC Netherlands, where she leads the HR Tech & Digital team within their People & Organization practice. 

With additional experience at tech giants LinkedIn and Microsoft, and as an author and sought-after speaker on HR Tech strategy and digital transformation, Marlene brings invaluable insights on how to successfully scale AI adoption from pilot to enterprise-wide implementation.

In this episode, you'll learn:

  • How to set up diverse and intelligent AI pilots
  • The key to scaling AI adoption across thousands of employees
  • Practical strategies for integrating AI into daily workflows
  • Methods for cultivating AI influencers within your organization
  • How to balance experimentation with strategy in AI implementation
  • PwC Netherlands' innovative "AI in the Loop" approach for ensuring fairness

Whether you're just starting your AI journey or looking to scale your existing initiatives, Marlene's experience leading PwC Netherlands' transformation will provide you with actionable insights you can apply in your organization.

Key Insights from Marlene de Koning

Here are the actionable key takeaways from the conversation:

1. Set up diverse and intelligent pilots

PwC began with 300 AI enthusiasts across various departments and seniority levels. This approach allowed them to test feasibility, usability, and effectiveness while considering data confidentiality. Create pilots that include people from across teams, departments, seniority, and AI mindsets to discover where AI can most benefit your organization.

2. Take a phased approach to scaling

PwC scaled from 300 to 2,000, then to 4,500, and eventually to all 6,000 employees. This phased approach allows for continuous learning and improvement. Involve HR or people teams in the process, as AI adoption requires workforce transformation. Plan your AI adoption in phases, allowing for feedback and adjustments between each stage.

3. Integrate AI into daily workflows

To combat the risk of people using AI once and then stopping due to unmet expectations, PwC focused on making AI tools like Copilot an integral part of employees' daily work. Practical efforts include:

  • Integrating AI into commonly used tools like Teams, Excel, and PowerPoint
  • Encouraging middle managers to remind and enable AI use in their teams
  • Highlighting best practices through user-submitted use cases and prompt libraries
  • Making AI learning playful through games like AI-themed Escape Rooms

4. Cultivate AI influencers

PwC identified natural influencers within the organization using organizational network analysis. To find and nurture influencers:

  • Use people analytics to identify the most connected individuals
  • Hold office hours where influencers can answer questions
  • Publicly celebrate and recognize influencers' contributions
  • Engage them at every step of the AI adoption journey

5. Balance experimentation and strategy

Marlene emphasized that successful AI implementation requires both a clear strategy and willingness to learn through hands-on experimentation. Develop a clear AI strategy for your organization, but remain flexible and open to learning from real-world experiences.

6. Implement "AI in the Loop" for fairness

PwC focuses on integrating AI into human workflows rather than the other way around. This approach helps address biases in AI models and ensures that AI supports human processes rather than replacing them. Consider bringing your company data into AI models to make them more bespoke to your organization's policy and culture.

By following these strategies, you can create a more effective and inclusive AI adoption process in your organization, leading to improved workflows and employee experiences.

Learn More and Lead with AI

Subscribe now to Lead with AI and be the first to hear my conversation with another AI leader who's transforming their organization.

Want to level up your AI skills today?

Want to get your whole team involved? Check out our offerings for customized AI courses for leadership teams.

If you have any other questions or feedback, or would like to be considered for the podcast, just send me an email: daan@flexos.work

🔔 Available on:

Transcript:

Daan van Rossum: You are now very famous, at least in my LinkedIn feeds, because, with PwC in the Netherlands, you ran a Microsoft Copilot adoption. Now, many companies, and especially many enterprises, are talking about AI and maybe thinking about AI, maybe even dreaming about AI. But you've actually done it. 

So I'm super curious about some practical lessons from that implementation. What was the original approach? What steps did you take, and maybe any lessons learned along the way? 

Marlene De Koning: Yeah. So first, to clarify, my role is focused on our external clients. So I run Copilot for our external clients, and our chief digital office is responsible for the Copilot rollout internally.

But because we'd like to learn from each other from what we do internally and externally, I'm also part of the AI steering committee and helping out in that regard. And we were one of the early adopters. So we talked with many of the partners directors about what course we're going to take.

And we started in November 2023, with around 300 AI enthusiasts and a pilot, and those were the experimenters to see what this product can bring us within the organization and to test the feasibility, usability, effectiveness, and what are the various scenarios where we could leverage Copilot, and what are the tasks that we could potentially take over from our people?

And then we decided to scale. So the first wave was around 2000 people; a few months later, mainly in our texts and advisory practice, we went to four and a half thousand people. And now, in September, the next wave will come, where we scaled even further. And if you think about our business, it makes a lot of sense because there's a lot of tasks where Copilots could be used in a very good way to make us more efficient.

Also, I think one of the things that we also see is that it's really reflecting on us that we were so early on because we did our annual people survey this year, especially amongst our junior colleagues or associate-level colleagues. They were very enthusiastic. They scored very high on engagement because of the innovation.

That's not only for Microsoft 365 Copilot but also because of the connection with OpenAI and the implementation of OpenAI. So I think we're going ahead of the curve, and that's being seen by our employees, and they love that they can bring that to our clients as well.

Daan van Rossum: That's incredible. So if there was ever a rallying cry for companies to adopt AI, if only it were for the fact that people in the organization see that as, okay, we're moving early, we're ahead of everyone else. That's exciting. It's exciting to be part of a company that's moving ahead and being ahead rather than following behind. 

There was a lot in there that you just shared. I was very curious about some of those steps. And I know that you're involved in another aspect of the AI rollouts, but how were those first 300 people picked? Because a lot of people are listening, they're going to be at different stages of AI adoption. Some may have experimented a little bit, maybe a little bit more advanced; what would that first step look like? How do you get to 300 people in an organization to say, Yes, I want to join on this AI journey?

Marlene De Koning: I think within many organizations, you might have more than 300 people who raised their hands and said, Yes, I want to be on a journey, especially in companies like ours, but I think there are a few things that are important when selecting different types of departments. So you can actually learn how it's been helpful within those certain departments. So that's definitely one of these.

So at the early stage, we were also thinking about the confidentiality levels of the data. So where will you use it, and where won't you use it? So that was also a reason to exclude certain departments within our organization from the first pilots. Because those were things that we still had to uncover to think about what our policies are and these types of things. Then you want to think about the levels. So you want to include leadership because, in the end, those also enable others to continue the usage and adoption. And then we looked a bit at the AI mindset.

So are there more experimenters, more innovators, and more agile users? So what type of users do you have? And we also looked at the businesses where people are. So how's that going to facilitate that? And also for organizations that want to start, these are definitely steps that you want to take into account as well, the data governance, because I think that's also one of the things that organizations tend to struggle with when scaling AI.

Daan van Rossum: Okay. So we definitely have to talk about the data privacy and ethics part of it. But what you just mentioned in terms of you basically had four criteria to look at. So how many people work at PwC in the Netherlands? 

Marlene De Koning: Around 6,000. And we rolled it out. So globally, many more territories are implementing Copilot. The Netherlands was ahead of the curve or is still ahead of the curve with the scale. So that's what we like as well, of course. But we started with around 300 out of those approximately 6,000.

Daan van Rossum: Out of all the people who raised their hands and said, I want to be a part of it. You actually made the very intentional decision to look at it; we want to have representatives from different seniority levels, different roles in the business, and different levels of AI mindset. Even I thought that was a really interesting one. 

How did that come about? Is that a framework that PwC developed globally and that you adopted? Or is this something that was developed in the business here? 

Marlene De Koning: I think a lot is coming from our future chief digital officer, Marshall Jacobson. He did a lot of this in this capacity. And I think the AI mindset is coming to something that we also bring to our clients, which is more about assessing psychological behavior. What type of people do you have within the organization? And then, how can you leverage that when you're looking for influencers and scaling the AI within your organization?

Daan van Rossum: Maybe let's switch a bit more to your world. So now you have this kind of first-hand experience of how to set up a pilot and how to get the first people in and then eventually to that bigger group. 

When you're now talking to clients, where do they stand in their decision-making process about whether or not to adopt AI? Are they already experimenting, and are you able to help them on that trajectory? Can I follow some of the best practices from your own pilot? 

Marlene De Koning: Yeah, I think it definitely helps that we've experimented ourselves and also that we are a very high-level regulatory organization that also helps in the conversation with clients on how to mitigate these risks, how to think about more responsible AI, and these types of things.

But also, we took the leap of faith, right? Because not everything could be measured in the early stages. And I think we're getting to more measurements; the longer you're using, the more you can pivot and tweak a bit. But in the early stages, we took a leap of faith. And that is something that many organizations struggle with, right?

What is the business case? How is this going to add value? How can we prove that it brings value to our organization? So within the six-week program that we developed, that we bring to our clients, and it can be a little more or less dependent on what the organization wants, but it's all about what are the different types of use cases?

So, where is your business? Will it bring what type of value? So you can imagine that it is also different for commercial organizations than for the public sector. Is the AI intended to deliver higher productivity and directly influence your P&L? Or are you looking to reduce the work stress of your employees, reducing complex sick rates?

What is the intention? Why you're implementing AI within your organization is one of the first questions that you need to answer, and then you need to see what the business use cases are that it could be used for. And then, in the case of Copilot, we also do a measurement program within that. So we've developed a measurement framework where we both look, preferably, at quantitative and qualitative data for our clients.

Then you can go as in-depth and crazy as you want with measurements. So normally, we do a comparison study, but you can even do a crossover study. And then, of course, the time takes longer because it's more intensely dependent on the measurement framework you set. And then, of course, adoption is a very big part of it.

And I think that is different for different types of organizations, because if you have low data literacy and low digital literacy within your organization, the adoption is way harder and also dependent on the trust level that people have in AI. Because if the trust is low in AI and the communication is not great, then you can buy the best AI or build the best AI there is.

But if nobody is using it, then you won't reach those results. So how do you then bring them on an adoption journey? And that's where, again, the AI mindset comes in. Understanding what type of people you have within the organization can really help. And then you can do adoption games, and then you can do prompt classes and these types of things for individual effectiveness, but also equipping really the leadership in how they help the organization adopt this part of the program.

And then, I think, the last thing that is really a prerequisite is cybersecurity and governance. Because that's also preventing organizations from suffering reputational damage or other types of damage, right?

Daan van Rossum: Yes, we can just imagine all the bad things that can happen. 

When companies come to you, have they already decided that they want to implement AI and therefore want to work and say, We're going to roll out that six-week program that you mentioned? Or is it also typically to consult on whether AI even makes sense for them, maybe based on things like digital literacy and the particular goals that a company has? 

Marlene De Koning: So we do both, and we also have a big data and AI team and a cybersecurity team within PwC where we do the full AI strategy for the organization.

My team is mainly focused on the human side of it. So if you think about even other AI tools or about the AI strategy that organizations are setting, we're looking way more at the impact on people. So how do you measure that business value? Also, how do you do the adoption and change program?

And then, because Copilot is such a workforce transformation for the organization and adoption for all that sits within my team, if you think about the more general, way more complex, and different AI systems, we have different teams for that.

Daan van Rossum: It makes sense.

Marlene De Koning: Yeah. And then, to answer your question on whether companies come to us, they come to us in different types of maturity stages.

Some did a pilot, or there's a very enthusiastic IT team that worked with a company such as Microsoft and they did a pilot, and then, how do you go from experimentation to scale? And I think that's where the experimentation phase begins. If you have enough enthusiasts in the organization, that happens, but then how do you scale and how do you work with the board?

And then, on one hand, if you think about the research that we do, as the PwC CEO says, they need to innovate, and their company's being disrupted in the next three years to come. But then, how do they act upon that? And how does this then contribute to the ambitions and goals that they set? That's what they need help with. And that's what they oftentimes go to us for.

Daan van Rossum: Yeah, that makes a lot of sense. They may have done those initial pilots, and that was something that they could execute internally, but then it gets to the point where, okay, now that you've won over the people who were already won over, they're already excited, right? At least to some degree.

Now you have to reach out to other people. And that's similar to when you went from maybe 300 people to 2000 people, eventually to 6,000 people. 

What are some of the steps that you see there as being very effective and very impactful when it comes to getting more people on board beyond those initial enthusiasts?

Marlene De Koning: I think bringing those on board and understanding who the influences are means that people are actually going to use it and also continue to use it. Because if you go to a larger scale, a big risk is that if you don't have a great adoption setting, people will use it once the AI doesn't live up to expectations, and then people will stop.

And especially if you look at the current marketing initiatives around Gen AI, it promises the world. But then the product truth is not always there yet, or it's hard to get there. So it takes some skills and some practice in order to get the best results.

So how do you then bring people on that journey where they need to do it multiple times before they get a better result and become more eloquent in working with the AI? That's not necessarily something that comes naturally to many people, and that's what happens with other implementations of the technology as well, of course, and they can get stuck.

If you think about other technologies, we've been buying a lot of them and a lot of systems over the past years as organizations, but they didn't necessarily communicate with each other, and actually, they made the work of people more complex, which made it more frustrating.

And that's also why adoption sometimes stagnates within the organization. And AI could actually be an answer for this, especially if you think about things like how, within Copilot Studio, you can enable that it takes action for you.

So you might not have to go to a different type of system, but you can stay in your flow of work. Type something in teams that will then fill out a form for you, and the prompt will ask you a question back, and they will fill out the right fields in the backend system, which of course enables a more seamless employee experience for people.

But that requires, as an organization, that you set your strategy first. What type of AI are we going to implement? And how are they going to communicate with each other? And I think that if you're going from experiment to scale, one of the big things that you need to think about is: how do we make it more seamless for end users? And then how do we integrate that into a full strategy?

Daan van Rossum: It's usually then when companies find out, as I hear from many experts, that we don't actually really know what the workflows are of a typical employee, especially at scale. We don't really have any clear insight into, okay, this is typically where they spend their time. And therefore, we can speed up their work, make it more joyful, and make it better by implementing AI.

And now obviously, we've seen some integrations, like ServiceNow being integrated with Copilot, and we've seen the Workday integration. So we're starting to get to this point where, okay, maybe I type something in Teams and all kinds of things happen in the backend, and that's very beneficial to me. 

But, like you said, that still requires some education. Somehow, we've all been used to using software as something that's very predictable. That's very systematic. We know that if we push that button, this happens with predictive AI, even if that were the case. 

Now we're getting to generative AI. It's obviously the opposite, right? It's a completely open interface, and anything can happen. So what are some things that companies can do as they're thinking about that leap from maybe that first experiment to scaling to make sure that people—you mentioned prompting classes, for example—understand that this is a teammate and that you need to communicate it almost in human-like language? You need to not prompt, but delegate, right? You need to actually be clear about what you're trying to get out of it to get the right output. 

So what are some things that you're then putting in place with your clients or with your teams to make sure that people get to that maybe mindset shift in how to use AI?

Marlene De Koning: Yeah. So we have also developed a few games where people can play escape group-type games and things like that. Our organizations that work with us can implement that. And the people, through playing, learn how a copilot works or how a system works. So then you take off already the first barriers on where's that button? How do I click? (See our recommendations for the best generative AI courses.)

It can be as simple as that. The people don't know where to find it. And then the barrier becomes higher. The more it's becoming democratized, the more people think that everybody knows how to work with it, except them. So then the questions become harder.

So if you bring that, and then if you bring more learning into a day-to-day, and you have the team learning and also equipping the leadership in asking the questions, So, if you think about myself, I work with a lot of junior people within our organization who, where I review the work, and before we had Copilots or Harvey and some of the other tools that we used differently than I review now.

So now I ask if they have a question. If they also ask the same question to Copilot or to check PwC, which is our internal on-demand Q&A basically, where people can then soundboard first before they come to me, and then, before they submit a presentation to me, they can have it reviewed by Copilot to see if there are spelling mistakes, how to rephrase things, or how to do things differently.

And that really helps me because it saves me time because everything is way more complete before it comes to review. And I think it opens up also for our entire leadership, more time to do other things where they normally spend it on review.

But it asks something from you—to not just start reviewing but to go back. So it's a behavioral change that you need to remind yourself of every time, and the same goes for meeting summaries or all these things. If you forget to turn it on, then you don't have to summarize, right? And those are the little things that people forget in their day-to-day lives.

So you really want to make it an integrated part of the daily lives of people. And I think that's the hard part with all technologies. But the beauty of things like Copilot is in your workflow. So, it is in Excel. It is in PowerPoint. It is in Outlook. So, it is in all the tools that you use on a daily basis.

So, it makes it easier for people to do the same work but slightly different in the same look and feel in the same surroundings.

Daan van Rossum: There are some nudges to get you to adopt AI and make it a habit. So we've seen from the Microsoft data that power users get exponentially more benefit out of AI. So, we want people to use it, have many use cases for it, and use it very frequently.

I think the Moderna CEO said in that interview that, he's saying 20 times per day or more, that should be the benchmark. Obviously, it's already in your workflow. If you're using something like Copilot, maybe with Chat PwC or Harvey, that may not be the case. 

What are some things that you can put into place so that people actually make it a habit that they all become power users?

Marlene De Koning: We did a few things, especially in the beginning; we had a weekly winner, which was basically where they submitted cases on how they used Copilot. And in the beginning, those cases were less sophisticated, right? Because it was the first week. And then we had winners, and they could win just a few things, but it was more for enthusiasts, so we shared all those prompts and the outcomes with other people.

So they learned what others were doing, how they were successful, and what it brought to them. And these became more sophisticated, and in the business unit that I'm part of, we have been fully deployed. So, we now move to a monthly sequence, but we're also having a prompt library, and we have a PwC global prompt library.

We share, across territories, what is working for us and how people could benefit from it. And then also thinking through the daily work that people do, and what are the tasks within that daily work that you can automate?

So, for example, we do a lot of workshops for our clients, and then you come home and you have a lot of outputs from those workshops, right? So all the notes, all the transcripts, even. And it needs to be bucketed. And it needs to be a write-up that goes to clients where you can actually have those workshops summarized, and you can leverage AI through that process in making that detailed output to your clients, which saves us a lot of power.

Think through those simple things that you do on a daily basis. And also, we do many surveys where we need to analyze all the open comments. Actually, that's also something where AI can save us a lot of work. So we thought through what our day-to-day uses were, and then maybe we did it for an employee experience survey.

But then, if you send it out to everybody, like we did for that one, the majority of the people in our organization are very smart. So then they think, Hey, we do other types of surveys, but we also do open comment analysis. So how can we leverage them? The same type of thing for it, right? So you create enthusiasm and a buzz.

And I think one other thing is to change agents or influencers—these types of things. So those are within the organization. So, we have them on all levels as well. Then, they are influential within their sphere as well. And it's not only coming from the top down that's very important; at least what we see as well.

Daan van Rossum: I see a couple of pillars there in terms of: it's not only looking at people's individual workflows but also looking at the type of work we do as an organization; there's a lot of surveys and a lot of workshops that need to be recreated and analyzed so we can share some common best practices.

Then again, once you see that as an incentive, it's really the carrot, not the stick, right? It's okay; of course I want to do that work quicker. Of course, I don't want to do that manually. So then eventually you're going to go over and then this idea of tapping also into leaders to say coach or motivate your people to submit work to AI before they submit it to you, right?

That's a win-win for both because they get immediate feedback rather than whenever you get around to it, and you get better work submitted to you once it's been through a couple of rounds of maybe copilots. And then finally, this idea of the influences, which I think is really interesting,. 

So maybe we can spend a few minutes on it. How do you create these influences, how do you find them, and how do you cultivate them? How do you make sure that they stay enthusiastic even if they get the 20th question about, well, I know you understand AI, and how do I do this? How do I do that? 

How do you make sure that they have the time for that, the energy for that, or the continued enthusiasm for AI?

Marlene De Koning: I think the last part you can easily mitigate by making sure that you have virtual office hours for these types of things that rotate between these people. So you take away a little bit of these, of that type of burden that might not be as fun.

But I think finding the right people is the more interesting part in this case. And that's not true; an organization suits different types of approaches.

Now, I'm also leading our people analytics team and PwC. So we love to do more organizational network analysis and then do a data analysis on who the actual influencers are, which is a more in-depth analysis and takes more time and resources to do, but then you can actually find who is in most contact with whom, right? And how much, and then you can really identify those influences within your organizations. And if you can put those people as change agents, then you see a huge uptake in adoption, regardless of technologies or what type of things, because they just influence people in a natural way. And there are natural influencers without ever being asked.

So if you didn't ask them to do something, that can be magical for your organization, but normally we ask who the influencers are or who is enthusiastic. And then I think it's also to the benefit of the organization to recognize the work that people do for the organization. And to put them in the spotlight or to thank them for the work that they do next to their work and that it doesn't become a commonality because people love normally what they do, but they like to be called out on it as well in a positive way.

Daan van Rossum: So a bit of recognition goes a long way toward cultivating the influencers and keeping them excited about AI. I can imagine that at some point it's going to be a bit of a burn as well. When you constantly get asked about these, maybe too simple questions, but keeping them excited and putting them in the spotlight, it sounds like that's a thing that helps a lot as well. 

Marlene De Koning: Yeah. And to counter that, I think what I see that our change agents are also getting difficult questions about is that the more they are challenging and that they can innovate, or that they never understood what their colleague was doing in their workflow, and that you now all of a sudden understand what a team workflow is and how a Copilot studio can actually help in part of the task automation. And you can equip a full team. So there are also these types of questions that are coming, which are a bit more challenging and hopefully more than burdening someone to balance it out, I would say.

Daan van Rossum: That's amazing.

There's definitely one question I want to ask you before we close out, which is that you wrote a book called HR Tech Strategy. And one of the core pillars of the book really revolves around this idea of fairness. And now we've spoken a lot about AI and all the adoption practices. I just had Helen Kupp Lee and Nichole Sterling from Women Define AI a couple of weeks ago.

We also talk about the fact that the data shows that women are taking up AI less than men, and somehow our community is quite balanced, but, broadly, the data shows that women are taking it up less than men. And then we have the additional difficulty that the AI is typically designed, written, and compiled by men.

And therefore, there is some bias inherent in the model. One of the tips that they gave is to make sure that, at least at a company level, when you're putting together, let's say this group of 300 or 2000, that you also look for diversity and that you make sure that everyone is included and that there is a diverse set of opinions and ideas around AI. What are some things you've put into place, as clearly this topic is near and dear to your heart as well? 

Marlene De Koning: Yeah, I think in PwC, we talk a lot about the responsible, so the ethical and the moral part of the framework. But also, we recently launched a paper, I think last week, around trustworthiness versus trust in AI and that we don't talk.

So what you hear a lot of other organizations talk about is the human in the loop. But actually, we started talking about the AI in the loop.

Daan van Rossum: Oh, I love that. Explain that's, that sounds very fascinating. 

Marlene De Koning: I think it is. And it is where, normally, you want to integrate the AI into human processes.

So if you think about what we're doing, we have tasks that we want to automate with AI. So that means that the human is always in the lead and running it where the AI comes in the loop, wherever it's necessary or wherever it adds value to the human. And I think that's a different type of mindset when you're talking about the implementation of AI, one that AI sees as more fit because it also makes you think more as an organization.

How does this increase the employee experience? How do we help and enable our employees? Because basically, we're doing this to make our lives better, or that's why we should be doing this. So if you can bring that more into it, and then of course you always have those risks of what you were just referring to the biases also in terms of culture,.

There are more trades on Anglo-American data. So it's owned by the bigger tech company. So those are all things that you want to take into consideration as an organization. What do I own? How can I potentially leverage the model but make it more tailored for my organization? So our policies and our culture can be fed into it.

And how can I make a more specialized model? That reflects more of what we as an organization need—questions that you can ask yourself and also work with.

Daan van Rossum: And that would allow you to rebalance maybe the current status quo and tip it more towards where you need it to be. I love that idea of the human in the loop versus the AI in the loop. And I'm sure a lot of people find that very reassuring—that idea that no, we still lead. We don't need to be in the loop; we lead. So I have three rapid-fire questions at the end of the interview. 

Number one is going to be your favorite AI tool. 

Marlene De Koning: Copilot and Teams.

Daan van Rossum: Second question. Your favorite AI tool besides Copilot and teams.

Marlene De Koning: Copilot and Teams!

Daan van Rossum: Okay. Is there a unique use case for AI in your daily routine? You've already talked about a bunch, but anything like in your daily work where you're looping AI in?

Marlene De Koning: I think in everything that can be automated, the ideal situation is to have the AI work really as a digital version of myself that takes away all the routine tasks and where I can focus on all the things that are more joyful for me.

So wherever I can implement AI, and yeah, it's in a safe environment, I try to implement that. So it is not a very unique case. I think it's just more mundane, and that's what I like.

Daan van Rossum: But a beautiful end vision is that eventually you will have a digital version of yourself that can take care of all the toil.

Finally, what's the one thing that any business leader should do tomorrow to get themselves and their teams onto the benefits of AI? 

Marlene De Koning: There's no such thing as one thing that organizations should do.

But if I need to narrow down, I would say to strategize and experiment. Please go hand in hand because you don't want to solely experiment without having a strategy, and you also don't want to set a strategy without ever starting the experiment. So I think those really go hand in hand, next to data governance, security, responsibility, and all the other things we talked about today.

Daan van Rossum: Sounds perfect. Okay, strategize and experiment.

Marlene, thanks so much for being on.

Marlene De Koning: Thank you, Daan.

Also available on:

Future Work - Listen on Spotify
Future Work - Listen on Apple Podcasts
Future Work - Watch on Youtube
TRANSCRIPT

Our latest articles

FlexOS helps you stay ahead in the future of work.

Atlassian #1 on Future 50 by Distributed Work

Atlassian #1 on Future 50 by Distributed Work

Atlassian has skyrocketed to the #1 spot on the Fortune Future 50 list, leaping from #26 last year!
Atlassian #1 on Future 50 by Distributed Work

Atlassian #1 on Future 50 by Distributed Work

Atlassian has skyrocketed to the #1 spot on the Fortune Future 50 list, leaping from #26 last year!
10 Themes for the Next Ten Years: Number 1 // Trillion Dollar Hashtag #2

10 Themes for the Next Ten Years: Number 1 // Trillion Dollar Hashtag #2

In 1981, Steve Jobs talked about how a computer was like a ‘Bicycle for the Mind.' But in these exponential times, ‘bicycle for the mind’ feels quaint, says Antony Slumbers. Read about agentic computing and what's next.
10 Themes for the Next Ten Years: Number 1 // Trillion Dollar Hashtag #2

10 Themes for the Next Ten Years: Number 1 // Trillion Dollar Hashtag #2

In 1981, Steve Jobs talked about how a computer was like a ‘Bicycle for the Mind.' But in these exponential times, ‘bicycle for the mind’ feels quaint, says Antony Slumbers. Read about agentic computing and what's next.
From Candidate Journey to Employee Advocates: Crafting a Lasting Employer Branding

From Candidate Journey to Employee Advocates: Crafting a Lasting Employer Branding

Unlock the secret to employer branding success: authenticity, AI, and employee advocacy for unbeatable talent attraction.
From Candidate Journey to Employee Advocates: Crafting a Lasting Employer Branding

From Candidate Journey to Employee Advocates: Crafting a Lasting Employer Branding

Unlock the secret to employer branding success: authenticity, AI, and employee advocacy for unbeatable talent attraction.
ElevenLabs Is Tranforming How We Create and Consume Content

ElevenLabs Is Tranforming How We Create and Consume Content

ElevenLabs’ text-to-podcast feature, Claude can now clone your writing, a tool to sync notes across devices, “David Mayer”, and more.
ElevenLabs Is Tranforming How We Create and Consume Content

ElevenLabs Is Tranforming How We Create and Consume Content

ElevenLabs’ text-to-podcast feature, Claude can now clone your writing, a tool to sync notes across devices, “David Mayer”, and more.
[Report] The World’s Most Popular AI Marketing Tools

[Report] The World’s Most Popular AI Marketing Tools

FlexOS.work surveyed AI platforms to reveal the leading AI Marketing Tools worldwide. Visual design and content assistants top the ranking, significant demand seen in Asia, and more insights for your adoption strategy.
[Report] The World’s Most Popular AI Marketing Tools

[Report] The World’s Most Popular AI Marketing Tools

FlexOS.work surveyed AI platforms to reveal the leading AI Marketing Tools worldwide. Visual design and content assistants top the ranking, significant demand seen in Asia, and more insights for your adoption strategy.
A GenAI Deployment Blueprint with Lessons Learned from Over 20 Pioneering Companies

A GenAI Deployment Blueprint with Lessons Learned from Over 20 Pioneering Companies

New global study from PwC x World Economic Forum highlights genAI’s potential with Case Studies and a Framework for Action. Also: breakthroughs in AI for image & sound, China’s model beats OpenAI’s o1, and more.
A GenAI Deployment Blueprint with Lessons Learned from Over 20 Pioneering Companies

A GenAI Deployment Blueprint with Lessons Learned from Over 20 Pioneering Companies

New global study from PwC x World Economic Forum highlights genAI’s potential with Case Studies and a Framework for Action. Also: breakthroughs in AI for image & sound, China’s model beats OpenAI’s o1, and more.
The Future of Workplace: 10 Themes for the next Ten Years

The Future of Workplace: 10 Themes for the next Ten Years

In the first edition of the Trillion Dollar Hashtag, real estate guru Antony Slumbers dives into predictions of the biggest themes for the future of work and real estate.
The Future of Workplace: 10 Themes for the next Ten Years

The Future of Workplace: 10 Themes for the next Ten Years

In the first edition of the Trillion Dollar Hashtag, real estate guru Antony Slumbers dives into predictions of the biggest themes for the future of work and real estate.
Do Your Employees Feel Seen and Valued?

Do Your Employees Feel Seen and Valued?

This week’s insights explore how gratitude, effective communication, and celebrating milestones (big and small) can drive meaningful impact in our workplaces.
Do Your Employees Feel Seen and Valued?

Do Your Employees Feel Seen and Valued?

This week’s insights explore how gratitude, effective communication, and celebrating milestones (big and small) can drive meaningful impact in our workplaces.