Speaking is Maartje Vennema, who helps organizations prioritize AI. “I help with the shift from can do to should do,” she explains. “What should organizations do to achieve their strategic goals?” That central question also features in her presentations, such as at Data Expo later this summer, and in the Chief AI Officer Academy she founded. “To start, I sit down with the executive team. Where are we headed? Next, we train an AI Officer from within the organization to identify and evaluate use cases, draw up a roadmap, and oversee the implementation.”

Lizzy Prins recognizes much of what her conversation partner is saying: “I help organizations get a handle on how AI is changing work, teams, and the organization. It’s actually already everywhere; there’s no shortage of pilots. But that doesn’t automatically mean you have an effective new way of working that contributes to business objectives.” Between a promising use case and actual organizational value lies a change challenge, Prins continues: “Work needs to be structured differently, responsibilities need to shift, and leaders need to make different choices. I help organizations with this through my company, High Potential Factory; I speak about it, and I’ve written a book on the subject.”

Sandbox experiment
Vennema also observes that many pilot projects are being run: “That’s great. I’m all for experimentation. But if you’re going to do that, let’s make sure we define when the experiment is a success—and when it isn’t. For larger projects, I believe you really need to have a clear objective defined in advance. What’s the payoff? If you don’t do that, it’s just a bit of playing around. Okay, fun, but now what?”

“I always call that a ‘sandbox experiment,’” Prins responds. “It might be a lot of fun, certainly enjoyable, but it should also contribute to something.” She notes that when organizations are asked to explain the intended added value, they often say that AI projects should save them time. “But what exactly is ‘more time’? What are you going to do with that time? What does it contribute to? These aren’t operational questions to be addressed after the fact, but strategic choices you have to make up front.”

Rushing into things without thinking it through
Prins observes that many organizations “rush into things without giving it much thought,” she continues. “AI projects are approved even though very few people—or really, no one—has a clear understanding of exactly what’s going to happen. For example, they assume that data quality is fine, which is usually not the case.”

“I recently saw someone on LinkedIn say that their organization’s data foundation wasn’t in order yet,” Vennema adds. “To which someone replied, ‘Oh, I’d use AI to get that sorted out.’ That’s pretty much the mindset we see a lot of these days. To some extent, AI can certainly help get the data foundation in order, but that’s not the whole story. These days, many people find AI and agents very sexy. They just find data boring.”

Prins: “If you know the foundation isn’t in order and you just throw AI on top of it, then everyone should really be scratching their heads. People do indeed see data as boring and tedious, perhaps because it feels like it takes time and energy. Meanwhile, AI has made the promise to the world that everything can be done in three seconds.” In unison: “That’s a shame.”

“AI promises that everything can be done in three seconds.”

AI is not a goal in itself
According to Vennema, executive boards must ask themselves three questions before giving the green light to an AI project. “First and foremost, it’s about the goal. And I’m not talking about an AI strategy, because that’s nonsense.” Vennema views AI as a means to an end. She continues: “What is our business strategy? That’s question number one. Second question: Can the idea we have contribute to that? If not, you might as well scrap the idea right away. And the third question is: What does it actually deliver? Is that enough to even get started? I think you can get very far just by asking these questions.”

“That’s exactly the problem,” says Prins: “Many organizations view AI as a standalone thing—a standalone project, a standalone strategy. But AI actually brings existing issues surrounding work, leadership, and organizational structure into sharp focus. That’s why you don’t need a standalone AI strategy, but rather clear decisions about where the organization wants to go and what that means for how people will work in the future.”

Critical thinking skills
According to Prins, AI is forcing organizations to think more than ever about added value. She therefore views the ability to assess added value as a key skill in the workplace, for both managers and employees: “We’re really going to work differently, and AI is a catalyst for that. Critical thinking is more important than ever before. These times call for the ability to take a fresh look at work: to determine what is valuable, where human judgment remains necessary, and how to reorganize work.”

“When it comes to the role of people in AI processes, you often hear that we all need to check the output,” Vennema notes. It’s called “human-in-the-loop.” “But it’s actually much more valuable to verify whether the thought process at the front end went correctly—that the algorithm is well-designed and made the right choices. Of course, that still requires substantive knowledge. I mean: an accountant can present a number that I can verify, but that still tells me nothing about his calculations and processes.”

“Of course, it’s helpful to stay in touch with the subject matter,” says Prins. “It will be a huge challenge for the people entering the job market now. They’re starting out with less experience under their belts, but of course they have every opportunity to gain substantive knowledge on their own to deepen their understanding. We’ll see where we stand in ten years—it will surely be very different from where we are now.”

Chief AI Officer
Prins questions whether the role of Chief AI Officer is yet commonplace within organizations. “I don’t really see it as a full-fledged C-level position. I’d actually find it more logical for those responsibilities to be spread across different people.”

Vennema, who, incidentally, is the founder of the Chief AI Officer Academy, says: “You might think I’m advocating for the same approach as the CTO and the CFO, but I’m not. I don’t necessarily see the Chief AI Officer as a position, but rather as a role that someone fulfills. It’s entirely possible to have multiple Chief AI Officers within an organization—one for each business unit, for example. So one for sales, one for marketing, and so on. They might report to the Chief Information Officer or the Chief Data Officer, for instance. These need to be people who know what’s going on—meaning they have operational knowledge in addition to strategic knowledge. And who, therefore, have the freedom to come up with use cases, create roadmaps, and lead implementations.” Vennema believes this is something for the coming years: “And I also hope that we won’t need it anymore after that.”

“ AI and agents are sexy; people just find data boring.”

Enthusiasm for AI
Vennema and Prins generally observe that executive boards are less enthusiastic about AI than the people on the front lines. “At the operational level, people see the possibilities and want to move forward with tools and agents,” says Vennema. “Executive boards are often a bit more cautious—and rightly so, because it shouldn’t turn into a wild ride. What I find unfortunate, just like you do, is that executive boards keep emphasizing that their main goal with AI is to save time and become more productive. But AI can also help them become more competitive and build better things, which makes work more enjoyable for people. That’s a whole different story.”

“I see a lot of great things happening on the work floor that organizations at the executive management level can learn from,” says Prins. “You have to be careful not to get bogged down as a board in debating the big picture without making any fundamental changes, while the front lines have already jumped on the fast track. That train needs to be slowed down just a little, but not too much; if you brake too hard, a lot of energy is lost. Perhaps the challenge is to provide direction: what do we want to strengthen, what do we want to limit, and what does that mean for the work?”

Vennema concludes: “As long as organizations don’t start implementing things like a headless chicken.”

Lizzy Prins and Maartje Vennema will be speaking at Data Expo.