WHEN you ask people about their experiences using AI in their private lives, enthusiasm is the prevailing response. Everything becomes much faster. Planning the next vacation is easy, and texts can be written in no time.
(By the way, that does not apply to this article. I have been working on it for quite some timeJ)
When you ask people who use AI at work, the answers become more nuanced.
Those with deep contextual knowledge who use AI as a sparring partner recognize its advantages, but they also see its weaknesses. Ask people who work on the technological side, and it becomes clear that building the AI technology stack is complex and that security risks are increasing.
A successful AI transformation is now the responsibility of the entire top management team, not just the IT department.
A comeback of process reengineering is needed. Just as in the 1990s, workflows must once again be analyzed and rethought across the entire value chain.
But let’s not fool ourselves: this time, the process will be significantly more complex because the division of tasks will no longer involve humans alone.
In the future, a team will also work with AI and independently operating agents. This will significantly affect how tasks and roles are distributed.
Of course, AI can store more explicit knowledge than any human, works extremely quickly, and does not get sick. However, people possess valuable implicit knowledge gained through experience, as well as intuition and empathy. Without these ‘human skills,’ no organization can function.
We must therefore clearly define the following:
Where is human judgment needed, and where does using AI make sense, both in terms of the work itself and economically?
Where can AI agents work independently, and where must a human make the final decision?
And how can we ensure that the necessary skills to make these decisions are available throughout the organization?
This is already complex enough in itself-and it becomes even more difficult because AI is constantly evolving.
Imagine all of this happening in a traditional organization where everyone works within inflexible structures and fixed job descriptions. It is clearly defined which team is responsible for which tasks and what skills are required. There are hierarchies, corresponding authority levels, and clearly assigned responsibilities.
Then there is another human factor to consider. In well-functioning teams, there is a working environment where colleagues help each other and newcomers learn from experienced employees. Decisions are discussed, including with the manager. People simply work together and collaborate.
However, the use of AI reduces human interaction, as Anthropic has already observed. Less interaction, in turn, leads to less contextual understanding.
If even the smallest problems are immediately directed to AI and are less frequently thought through independently or discussed with colleagues, this does not help develop people’s general cognitive abilities!
I am not only concerned about younger people who are growing up with AI and whose neural connections are not yet as firmly established simply because of their age.
In my view, no one is immune to the temptation to take what seems to be the ‘easier’ route, even when faced with simple questions, rather than making the effort to think things through independently.
I am particularly concerned about organizations where continuous learning is not an established part of the company culture.
billboards do the job socmed
For these organizations, the AI transformation will become an enormous leadership challenge.
Throughout this journey, the message for managers is clear: coach more instead of controlling, build trust, and communicate openly. This includes being open about their own uncertainties and discussing how the roles of humans and AI are changing.
Here are the three most important lessons:
1. AI should support people, not replace human judgment. AI can process information quickly, but humans still contribute experience, intuition, empathy, and decision-making.
2. People must continue learning and thinking independently. If we ask AI to solve every small problem, we may lose opportunities to develop our own thinking skills and learn from colleagues.
3. Leadership must adapt to AI. Managers should help their teams understand which tasks belong to AI and which require human involvement. They should encourage collaboration, trust, and continuous learning.
In simple words: AI is a powerful tool, but successful companies will be those that combine AI’s capabilities with human intelligence, experience, and teamwork.