AI mindset is no longer about whether we use AI – Switzerland's 82% adoption rate among knowledge workers proves it has arrived. The real question is whether our AI mindset is ready to let it reshape how we work at the core. The divide is not between those who use AI and those who don't. It is between those who let AI touch the surface of their work and those willing to rethink the work itself. [1][2][3]
And yet, if I look honestly at my own work, I see a contradiction. I use AI regularly. Not as a gimmick, but as part of how I think and deliver. I use it to develop ideas for concepts, draft concept papers, summarise meetings, and discuss options when I want to challenge my own thinking. In that sense, AI is already embedded in my day to day work. [3]
But only up to a point. There are still parts of my work where the mindset suddenly changes. In those moments, AI is no longer treated as a real thinking partner but as an assistant at the edge. It can summarise, polish, or document. But it does not really shape the work itself. That is where this reflection starts for me. Not with technology, but with my own habits.
Where I already trust AI and where I do not
There are areas where AI already adds obvious value for me. I use AI to generate angles, structures and early draft versions for concepts instead of starting from a blank page. I use it as a conversation partner to test options and expose weak spots in my reasoning. And I use Copilot to turn long meetings and message threads into summaries I can actually work with. So this is not a story about someone who is sceptical of AI or late to adopt it.
But there are still clear limits in how far I let it in. PowerPoint is one example. I would gladly use AI much more for presentations, but the current integration is still so weak that the output often creates more work than it removes. The tool is there, but the quality is not. And then there is a second category that feels even more revealing.
In my own world, that is journey design and requirements gathering. But I do not think this pattern is unique to my role. It shows up anywhere work depends on coordination, judgement, synthesis and structured thinking. Strategy teams, product managers, project leads, compliance professionals, HR teams, business analysts. Many of us use AI in visible, low risk ways, while still protecting the core parts of work that define how we create value.
The AI Mindset Divide: Trust vs. Protection
This is where the Microsoft numbers become more interesting than they first appear. AI usage is already high, but that does not mean work itself has been redesigned around it. [2][1][3]
I can see that in myself. I am comfortable using AI to get to a first draft faster. I am comfortable using it to challenge an idea before I send it out. But when work becomes more ambiguous, more political, or more central to how decisions are shaped, I notice how quickly I return to traditional methods. Then it becomes meetings. Reviews. Iterations. Manual alignment. Familiar rituals.
Journey design and requirements gathering are simply the clearest example from my own environment. These activities are still often organised as if the best way to create clarity is to collect the right people in enough rooms for long enough. Inputs are gathered manually. Journeys are mapped manually. Requirements are drafted, reviewed, challenged, rewritten and circulated again.
AI may take notes in the background, but it rarely shapes the actual structure of the work. And that, I think, is the real divide. It is not between people who use AI and people who do not. It is between those who let AI touch the surface of their work and those who are willing to rethink the work itself. [3]
What could already be possible today
If I look at what is technically possible now, not in some distant future, our current way of working already feels conservative. In journey design, AI could help identify patterns across complaints, tickets, interaction logs and process data before a workshop even starts. Instead of beginning with opinions and memory, a team could begin with a proposed map of what is actually happening.
In requirements work, AI could help turn static documents into living artefacts. It could detect overlaps, inconsistencies and outdated assumptions across versions and propose updates when new signals appear. But this logic goes far beyond my own field.
- In strategy work, AI could generate competing narratives, surface second order effects and challenge hidden assumptions before a leadership discussion begins.
- In project management, it could identify recurring blockers across initiatives, flag coordination risks early and suggest where dependencies are likely to break.
- In compliance or risk, it could trace patterns across cases, cluster emerging themes and highlight where existing controls no longer fit current behaviour.
- In product and service design, it could continuously aggregate feedback, detect where customer friction is increasing and suggest where a journey or interaction should be rethought.
None of this removes the need for human judgement. But it does challenge the idea that judgement must always begin with manual synthesis.
Why Adoption still stalls
If so much is possible already, why do so many professionals still stop halfway. Part of it is tool quality. When visible AI features disappoint, it becomes easy to dismiss the broader potential as well. Part of it is organisational ambiguity. Many companies encourage AI usage in principle, but have not yet defined how AI should systematically change how work is done. [2][3]
But a large part of it is more personal than that. AI enters the very parts of work that many people associate with their professional value: structuring ambiguity, creating clarity, asking the right questions, writing the first narrative, connecting the dots. These are not side tasks. For many professionals, this is the job.
That is why adoption often stops at the edge.
- AI can summarise my meeting, but not reshape my process.
- AI can help draft my text, but not question how I produce decisions.
- AI can support my work, but not challenge the rituals around it.
That is a much more human barrier than most AI discussions admit.
Conclusion and the uncomfortable Question
So for me, this is no longer a question of whether I use AI. I do. The more uncomfortable question is this:
Where in my own work am I still defending a traditional way of working, not because it is better, but because it keeps me in familiar control
Journey design and requirements gathering are part of that answer for me. But they are only my version of it.
Everyone has their own protected zone. The part of the job where AI is allowed to assist, but not to fundamentally reshape how thinking happens.
And maybe that is the more provocative implication behind the Microsoft study. If AI adoption in Switzerland is already high, then the next divide will not be between people who use AI and people who do not. It will be between those who use AI to improve the edges and those who are willing to let it trigger a deeper reflection on how they work at the core. [1][2][3]
That is not mainly a technology question anymore. It is a question of mindset.
