What is AI?

There are many different ways to classify and sort AI. Sometimes we talk about seven categories, sometimes four. These categories often describe AI systems' capabilities from a technical perspective, ranging from simple reactive machines to what is called superintelligence or artificial general intelligence (AGI). In most cases, it’s more useful to understand AI in terms of three other categories:

Generative AI, which can create things that didn’t exist before, such as images or texts.

Analysing AI, systems that process large amounts of data to find patterns, create summaries, or detect anomalies.

Productive AI, systems that actually perform tasks for us, like driving cars, drafting and sending emails, regulating building temperatures, or constructing houses.

New ways of working and new organisations

AI enables, and sometimes necessitates, new ways of working. Most organisations base their work on three concepts we can call hierarchy, specialisation, and planning. These concepts have their roots in the early days of industrialism and were probably the best approach given the limited capacity for information management at the time. However, with nearly unlimited capacity for information handling, we can replace these three with their opposites whenever it benefits us. We can work non-hierarchically, with complexity, and reactively.

AI-enhanced operations

Different operations can develop, and will be impacted, in very different ways. In some areas, the discussion will focus primarily on the risk of fraud and deception, in others on entirely new professional roles, and in yet others on how much more we can achieve with far fewer resources. These discussions are not predetermined; we choose how we want to approach AI at an operational level. To reach that third level, where we talk about better operations and improved sustainability, here are the first three steps: Choose a pilot operation, start with analysing AI, engage all employees.

Quality and follow-up

Quality work can be developed with the help of AI in several different ways. It’s no exaggeration to say that nearly all quality work will be based on some form of AI and the use of vast amounts of data. This also enables "global" quality work, meaning that instead of being limited to the information and analyses within our own organisation, we can, in many cases, use data from entire industries to understand what is happening—and what is not. What is genuinely new, compared to all previous quality work, includes the following: We can use all operational data as a basis. We can identify quality issues as they occur, or even in advance. We can ask AI to search for correlations.

Users and customers

One effect of AI concerning users/customers/members or citizens is that their expectations of our services and operations will increase. Another is that we can have a completely different level of communication and relationship with them, and a third is that their ability to "self-serve" can also increase. Those who move the fastest will set the benchmark. Users and customers will become an integrated part of our operations. Empowerment and self-determination.

Competence Supply

To begin with, our understanding of what competence is—and perhaps more importantly, where it resides—is being challenged. For most, competence supply is a concern for HR, as competence is something that people have (or lack). What AI introduces into the equation is that a person who lacks competence might become competent with the right technical support. Another aspect is that AI is almost certainly the greatest revolution in human learning since Gutenberg, and probably even bigger than the printing press. Employees with AI support can become nearly omnipotent. Employees will need a different type of competence. AI-supported learning is a completely new phenomenon, and we also need to learn more about it.

Security and Privacy

Our operations can be greatly improved with the help of AI, but there are also significant risks to manage. Some risks may seem so large that we might be tempted to avoid the technology altogether. However, that is likely not a viable path. Instead, these risks must be identified and, where possible, mitigated—or the organisation must be prepared for potential incidents stemming from AI-related problems.

Strategy

Right now, we are all likely in a phase of being a bit delightfully (and somewhat alarmingly) caught off guard. We are experimenting, more tentatively than purposefully. That’s probably necessary—but amidst this uncertainty, we also need to gradually shape a strategy that provides a stable foundation for long-term organisational development, technological infrastructure, and competence supply, while also preparing for the next leap—a preparedness very few of us had this time.

We need one (1) strategy. It is no longer effective to formulate separate strategies for areas like HR, finance, product development, or digitalisation. The organisation needs to be more flexible. The entire strategy must be embodied throughout the organisation..

How Do We Learn to Live in an AI-Powered World?

Those who wait until AI is good enough for “real operations” or safe enough for handling personal data and trade secrets will fall behind. This technology is advancing so rapidly that the only realistic approach is to experiment with it—using harmless tasks and processes. Through this, we learn how to use it and what it does to organisations and professions. That way, we’ll be ready when the technology moves beyond the prototype stage—which will happen in a year or two.

Digital twins

A truly exciting development is the possibility of creating a digital twin, a simulation of an organisation in a computer. In the digital twin, various scenarios can be simulated, but perhaps most interestingly, it allows us to test whether our own assumptions about how the organisation functions are actually true and relevant.

Workshop: Create a Travel Agency

In groups of five to seven people, we’ll create an imaginary travel agency with a new business idea and a new target group. We’ll use AI for every step, from marketing and logo design to system development and business planning.

Workshop: Explore What's Available

We’ll look at how AI can currently be used, including language models, plug-ins in Office applications, image generation, and programming. The workshop can be conducted with or without participant activity, depending on the technical conditions.

Workshop - Thing

My AI tool, Thing, listens to the conversation in the room, summarises, comments, and reflects in real-time—while also providing meeting notes and a complete transcription of what has been said. It can be used as a standalone workshop or as an accessory in other contexts.