Leading up to his keynote, we spoke with Frans about the opportunities as well as the challenges presented by today's abundance of data and the rise of AI.
Data-driven innovation
At first glance, it seems like a dream scenario: unprecedented amounts of data, with which you can make better decisions, get to know your customer better and develop and deliver new services. Practice paints a different picture. Research shows that 70 to 80 percent of data projects fail to achieve their intended goals. "Those are obviously worrisome numbers," says Feldberg. "We have the technology, we have the data, but we apparently fail to fill in the necessary prerequisites for success."
So how do we translate the concept of data-driven innovation into practical and successful implementation? Both in business, where business models are under pressure from digitization, and in the public sector, where data can help with social issues such as healthcare, security and sustainability. According to Feldberg, the key to success lies in developing a common story around data- and AI-driven innovation. A story to which all relevant stakeholders, business, IT and data professionals alike, contribute. "When it comes to AI, people are quick to shout: we have to do something with AI! But why start with the 'how'? Isn't the 'why' more important? And, do you have a clear understanding of what AI can and cannot do? These are questions that need to be answered in such a story.
Data requires customization
Data and AI are much more than technological developments. According to Feldberg, they touch every element of the business model, or task in the public sector. From how an organization creates value, to how it delivers that value to its target audience and how it manages to retain it internally.
"At its core, a business model is about three questions," he explains. "How do you create value? How do you deliver it to the right target audience? And how do you ensure that value remains at the bottom line? Data and AI potentially affect all those aspects."
Companies like Facebook, but also more and more innovative startups, have an unprecedentedly detailed customer view, often sharper than many organizations themselves have of their own customers. Feldberg: "That confronts you as an organization with an important question: how do you stay relevant when someone else knows your customer better than you know yourself?" And it goes beyond customer relationships. Data and AI are also changing the relationships between people, businesses and governments. They influence internal processes, external collaborations and the expectations that customers, citizens and employees have of organizations and each other.
That confronts you as an organization with an important question: how do you stay relevant when someone else knows your customer better than you know yourself?"
"That raises strategic questions," says Feldberg. "What does this mean for your relationships with your customers? With your employees? Your suppliers? Your business partners? And for collaboration between departments or teams within your organization?"
The tricky thing, and at the same time interesting, is that there is no blueprint for what the impact of data and AI will be. No universal answer or standard approach. Each organization operates in its own context, with its own data, dynamics and goals. And so it requires customization. To learn and dare to experiment. But, it also offers opportunities to differentiate yourself as an organization, both public and private.
A shared story is essential
Anyone who thinks the success of data and AI projects depends primarily on advanced technology will be disappointed. "People often think technology is important," says Feldberg, "but it is rarely the deciding factor in whether or not a data project succeeds." Extensive research on critical success factors shows that it is an interplay between organization, people, governance, data management and technology.
Developing a shared story is the solution. It can be that simple."
"Data projects are by definition multidisciplinary," he explains. "You need business professionals, IT professionals and data professionals - and they often speak very different languages. If you don't make sure these people understand each other, share knowledge, communicate well with each other and develop a shared story together, it's almost impossible to make such a project a success."
Research has shown time and again that this process, called "alignment", is one of the most important prerequisites for success. "You can have the best infrastructure and the smartest algorithms," says Feldberg, "But if the different teams don't understand each other, then success is very far away. Developing a shared story is the solution. It can be that simple."
Beyond the hype
Is AI fundamentally different from previous data projects? "Yes and no," says Feldberg. "At its core, artificial intelligence is a data-driven technology. So many of the success and failure factors that apply to data projects also apply to AI projects. Data quality, for example, not sexy, but essential."
The rise of generative AI (such as ChatGPT) has captured the imagination of the general public and business world. The speed at which new applications are emerging is unprecedented. This presents a new challenge for organizations: do you have the absorptive capacity to move with them, to understand what is relevant and what is not? In this, understanding what AI can and cannot do is essential. How else can you judge what to do and what certainly not to do? And how to do this responsibly as well.
According to Feldberg, there is a great risk that organizations will get lost in the noise and make decisions based on incorrect expectations. "The messaging around AI is often 'glaring' and unsubtle: AI will take over your job, computers will become smarter than humans, within now and two years everything will be different. AI is presented as the Haarlemmer's oil for all problems. But it certainly isn't!"
To assess all these claims, Feldberg says it is very important to know what types of AI there are and what their limitations are. "If you know this, you can much better assess where the power of AI lies and deploy it successfully and responsibly." He continues: "For example, AI cannot think. Thinking includes being able to estimate the purpose of an object or human being without being clear beforehand. This is called 'intentionality.' For now, this is one of the biggest challenges for AI. And so there are more. However, those who blindly follow the hype often create unrealistic expectations. Then the wrong discussions arise and disappointment follows. Not because AI does not work, but because it is not clear beforehand what AI can and cannot do for your organization." In his keynote, Feldberg elaborates on this. Because, as the great scientist and philosopher Goethe put it in the 18th century, "In limitation the master shows himself.
Photography: Ruben May
In addition to his work as a professor and researcher, Frans is a sought-after speaker. He has appeared on numerous stages at home and abroad, inspiring organizations with his clear perspective on data and technology. During his keynote, Feldberg discusses the different types of AI, their limitations, and how to deal with them effectively. He does this based on theoretical insights that are explained in a low-threshold manner using inspiring examples and appealing cases. Feldberg: "Of course I also offer concrete tools for the successful and responsible use of AI in your organization. So the question, "What can we start with tomorrow?" will also be answered. Because one thing is certain, the opportunities offered by AI must be utilized to stay relevant! So after my keynote you can call yourself 'master'."
Frans Feldberg, Professor at the Free University of Amsterdam, will open Data Expo on Thursday, September 11, with the keynote "AI: in the limitation the master shows himself!" Order your free tickets here.
