Not long ago, some of the least glamorous work in an organization served an important purpose: it was the work that got you started.
A new employee prepared the first draft of a slide deck. A junior team member gathered the information, checked the numbers, or sat quietly in a meeting taking notes. The work was rarely exciting. It was not always efficient.
But it was where people learned.
They began to notice which questions mattered. They saw how experienced leaders evaluated competing ideas. They learned the history behind decisions and the difference between an answer that looked right and one that made sense in context.
Today, much of that work can be completed by AI in seconds. That can feel like progress. In many ways, it is. But something else may be disappearing with every task delegated to the ghost in the machine.
The Experience Hidden Inside the Work
Gartner recently warned that AI could eliminate many of the traditional entry-level responsibilities that helped prepare people for leadership. A Gartner survey of 1,303 senior leaders found that:
- 18% of marketing leaders had eliminated roles because of automation
- 16% had created new roles
- Nearly one-third had redesigned roles
The concern is not simply that certain jobs may disappear. It is that the experiences inside those jobs may disappear with them.
When someone prepares an analysis, they learn how the pieces fit together. When they attend a difficult meeting, they see how influence actually works. When they produce an imperfect first draft, they receive feedback that helps them recognize what good judgment looks like.
These moments build more than technical skill. They build context, discernment, and confidence. Those qualities are difficult to acquire from a course. They grow slowly through exposure, responsibility, mistakes, observation, and conversation.
In other words, they grow from experience, and the opportunity to experience may be in jeopardy.
Efficiency Has a Longer Shadow
A leader looking at a task may reasonably ask:
“Can AI do this faster?”
That is an important question, but it may no longer be enough.
There is another question beside it:
“What was this work teaching the person who used to do it?”
A task can appear routine while also carrying significant developmental value. Remove enough of those tasks, and an organization may become more efficient today while quietly weakening its leadership pipeline for tomorrow.
People may advance without having seen how decisions are formed. They may know how to produce an answer without understanding how to evaluate it. They may enter management without the organizational knowledge that once accumulated through years of ordinary work.
The gap may not be visible immediately. It may only become apparent when the organization needs someone ready to step forward.
A New Responsibility for Leaders
This creates a different kind of leadership challenge.
The goal is not to preserve unnecessary work simply because it is familiar. Nor is it to resist technology that can free people from repetitive tasks. The opportunity is to become more intentional about where development will happen.
If AI takes away one path for learning, another path has to be created. That might mean inviting a less-experienced employee into a strategic conversation. It might mean asking someone to evaluate an AI-generated recommendation instead of merely accepting it. It might mean giving people responsibility earlier, along with the guidance and room to make mistakes that responsibility requires.
The future of leadership development may depend less on formal programs and more on the choices leaders make about everyday work. I’d go so far as to say that this is has always been true. Every assignment communicates something about who is trusted to think, contribute, and grow. As AI changes what work requires, it may be worth pausing before removing a task entirely. Today’s leaders have to ask themselves if the way it used to work is worth preserving, at least in some form, so tomorrow’s leaders have the opportunity to build the critical skills to make themselves successful.
Unless, of course, the goal is to let AI take the lead.
What might we be automating besides the work?
