AI is beginning to change the work inside familiar roles.
Research that once took several hours may take minutes. An AI system can prepare the first analysis or produce a draft before a junior employee has learned how to build one. Routine cases can move to AI while the remaining work requires more interpretation from the person responsible for it.
The job title may remain the same while the substance of the work changes underneath it.
The International Labour Organization expects much of AI’s near-term effect to come through changes in tasks and the organization of work, rather than the removal of entire occupations. That makes one question especially important for leaders:ilo
Before we change the work, how well do we understand what doing that work has already taught our people?
Most organizations can describe the skills required for a role. They know where employees have worked, which responsibilities they have held, and how their performance has been assessed.
What is much harder to see is how years of work have changed the way a person thinks.
Experience changes more than skills
We usually describe professional development through skills because skills are relatively easy to name.
Someone learns to manage a project, analyze financial information, or lead a customer relationship. With experience, they become more capable at the work.
But years of doing something also change how a person approaches it.
An operations leader who has spent a long time inside complex systems may notice the early signs of a process beginning to fail before the data provides a clean explanation. That knowledge is not mystical, and it is not simply intuition. It has often been built through situations that went wrong, decisions that had unexpected consequences, and enough repetition to recognize what tends to happen next.
A manager may learn that a performance problem is rooted in the design of the work rather than in the person doing it. Someone who has spent years leading change may become better at distinguishing between temporary discomfort and a deeper problem that requires attention.
I think of this as professional knowledge.
It is more than the ability to perform a task. It includes the understanding that develops around the work and changes how a person uses what they know.
Experience alone does not guarantee that kind of development. People can repeat the same habits for years without examining them very closely.
What matters is whether something in the experience changed their judgment.
Expertise becomes difficult to see
One of the paradoxes of expertise is that it can become less visible to the person who has it.
A lesson that once demanded deliberate thought eventually becomes part of the way someone works. They no longer experience it as something they learned because they have used it so many times that it feels ordinary.
I have seen this repeatedly with experienced professionals.
Ask someone what they are good at and the first answer may sound much like their job description. Stay in the conversation a little longer and a different picture begins to appear.
They start talking about the problems colleagues bring to them when the obvious answer has not worked. They remember a situation that permanently changed the way they handle something. Sometimes they describe a pattern they can recognize almost immediately today, even though it would have taken them much longer to see earlier in their career.
Those are often better clues to professional knowledge than the familiar adjectives people use to describe themselves.
They also reveal something an organization chart cannot show: where people turn when a problem no longer fits the standard process.
OECD work on knowledge transfer makes a similar distinction. Formal information can be documented and transferred relatively easily. Tacit knowledge remains with experienced employees and in an organization’s collective memory, making it harder to articulate and share.oecd
That difficulty does not make the knowledge less valuable. It makes it easier to overlook.
I saw this first in operations
I spent more than two decades in operations. For most of that time, I thought about experience through the work in front of me.
There was always another operational problem requiring attention, another decision to make, or something in the organization that needed to work better. I did not have a separate place where I recorded what those situations were teaching me.
I first saw the consequences from the employer’s side.
An experienced person would leave. Before long, the organization would discover that part of what they knew had never really existed in a system. Other people could eventually learn it, but sometimes the company spent months rediscovering something that had already been understood.
Later, when my own career began moving in a different direction, I encountered the same issue from the other side.
I could describe my roles easily because that language was familiar. I knew how to explain the responsibilities and the work I had led.
It was much harder to separate the marine industry from what more than twenty years of complex operations had taught me about people and organizations.
Some of that knowledge belonged specifically to the environment where I learned it. Other parts were much more portable than I had understood at the time.
It took reflection to see the difference.
That experience changed the way I think about the knowledge held across an organization. A company may have detailed information about roles and formal skills while knowing far less about what years of work have taught the people performing them.
What changes when AI enters the work
When an AI system takes on a task, leaders naturally look at what becomes faster.
That is an important part of the decision, but it is not the whole decision.
The task may also have been giving employees repeated contact with the underlying problem. It may have taught them which information was usually incomplete, where assumptions tended to fail, or when a technically correct answer did not fit the situation.
If AI prepares the first analysis, the person may enter the work later than they did before. If routine cases are automated, employees may encounter fewer of the situations through which they once learned the profession.
This does not mean inefficient work should be preserved merely because it is familiar.
It does mean leaders should understand what the work was teaching before they remove or substantially change it.
The same question applies to experienced professionals.
AI may reduce the value of a particular method without reducing the value of everything a person learned by using it. Someone who spent years producing sophisticated reports may know far more than how to produce the document. They may understand which questions need to be asked before the report can be trusted.
If leaders see only the task, they may miss the knowledge built around it.
A better view of the workforce
Most workforce systems were not designed to show this kind of knowledge.
They record employment history, roles, performance, qualifications, and skills. Those records serve important purposes, but they reduce the complexity of experience so that the organization can make decisions efficiently.
The harder knowledge often appears elsewhere.
It appears in the difficult decisions someone has made. It appears in the problems colleagues repeatedly ask them to solve. It can be found in the project that failed and permanently changed how they prepare for the next one.
Leaders do not need to document every lesson held by every employee. Nor should an organization assume ownership of everything a person has learned during their career.
But leaders do need a better way to notice where important knowledge sits.
That begins with better questions.
When managers discuss development, they can ask what recent work changed in the way the person thinks. When a role is redesigned, they can examine which tasks merely consume time and which have helped people develop the judgment the organization still needs.
When an experienced employee moves on, the conversation should extend beyond a handover of files and current responsibilities. It should include the recurring problems they learned to recognize and the decisions that require more context than a written process can provide.
These conversations give leaders a clearer view of the workforce than a list of skills alone.
Before redesigning the work
Much of the conversation about AI and work focuses on what people will need to learn next.
That matters. The work is changing, and people will need opportunities to develop with it.
But an experienced workforce is not beginning from zero.
Years of work have already produced knowledge that may not appear in a job description, a skills profile, or an organizational system. Some of that knowledge will become less relevant as methods change. Some may become more important when AI carries out more of the routine work and people are responsible for interpreting the result.
Leaders need to know the difference.
Before redesigning the work, they need to understand what doing that work has already taught their people.
