If artificial intelligence takes on Junior tasks… how do we build Senior expertise?
Listen to this article
Read by Anchor
In the past we sometimes saw routine tasks as just work that had to be finished before the employee could start doing “the important work.” Over time I began to see it differently.
When a new employee spends two hours reviewing a report, they are not only correcting errors. They also start to understand what a good report looks like, notice recurring mistakes, and learn which details matter and which can be ignored.
And when they write a draft and their manager returns it with ten comments, the experience may be frustrating at the time, but it is a very important part of building expertise. When they analyze the numbers themselves and later discover they misunderstood them, the mistake itself adds something for them.
In other words, the task produced work… but at the same time it produced experience.
And that is the part I fear we will forget as we try to achieve the highest possible efficiency with artificial intelligence.
Experience does not come from the correct answer alone
The expert employee is not the person with the greatest amount of information. If the matter is only information, artificial intelligence will probably know more than any of us.
In my view, expertise is something deeper.
- It is being able to see a situation and realize that an answer that looks logical on paper is not necessarily the best here.
- Knowing when you need additional information and when you must make a decision.
- Seeing a problem that occurred before and saying, “I’ve seen something like this.”
Expertise is built from very small situations that accumulate: attempt, mistake, feedback, retry, decision, then seeing the decision’s outcome.
That is why we can read ten books on leadership, yet that will not make us experienced leaders without actually leading people. We can study dozens of decision-making cases, but good judgment forms when we make real decisions and live their outcomes.
Artificial intelligence can accelerate access to information, but it cannot compress five years of professional experience into five minutes.
A situation that made me think about the issue differently
In one work situation I was reviewing an output prepared by a colleague using artificial intelligence. At first glance the result was very good: well-structured, its language excellent, and the ideas logical.
But when we began discussing it, I discovered that the real value was not in the output’s quality itself. The value lay in the questions:
Why did we choose this suggestion? Does it truly fit the context we work in? What could happen if we implement it? Which part do you agree with and which part are you not convinced about?
At that moment I thought the new challenge might not be that the employee knows how to use artificial intelligence to produce a good answer.
The challenge is for them to have enough experience to know whether the answer is good in the first place. That is a major difference.
The role of managers and L&D may need to change
If part of learning used to happen naturally while doing the work, and with artificial intelligence some of those tasks have started to disappear or shrink, then we need to deliberately compensate for that.
In my view this is a shared responsibility between managers and learning-and-development specialists. It is not enough to teach the employee how to use artificial-intelligence tools.
- We also need to design opportunities for them to build their professional judgment.
- Let them try to solve the problem before seeing the answer.
- Ask them about their reasoning, not just the result they arrived at.
- Give them real scenarios, a space to make a decision, and feedback after the decision.
- And sometimes encourage them to use artificial intelligence not to “give them the solution,” but to challenge the solution they have thought of.
For example, instead of asking, “What is the best decision in this situation?”
They could ask, “This is the decision I consider appropriate. What risks might I be overlooking?”
Here artificial intelligence becomes a tool to expand thinking, not a substitute for it.
We don’t have to choose between efficiency and learning
I certainly don’t see the solution as sending employees back to do everything manually so they “learn the hard way.” That runs counter to the nature of progress. If a tool saves three hours of work, it is natural to use it.
But we may need to ask an additional question before we fully automate any task:
Is this merely a task… or is it also a learning opportunity?
Because there is a difference between eliminating a repetitive, valueless task and eliminating an experience that helped the employee build a skill they will need later.
- Artificial intelligence can write the first draft, but periodically let them write it themselves.
- It can analyze the data, but let them explain what they see in the numbers.
- It can propose five solutions, but ask the employee: which one would you choose and why?
Efficiency matters. But building people matters too.
In the end
I believe we are entering a new challenge in talent development. For years we have been thinking about how to make the new employee learn faster.
Now we need to add a tougher question: how do we let them benefit from the speed of artificial intelligence… without shortcutting the experiences that build their expertise?
Because the problem is not that artificial intelligence will perform some junior tasks.
The problem will begin if it does everything, and we have not thought of a new way to build Senior expertise.
In my view, this could become one of the most important roles for learning and development in the coming years: not just teaching people how to use artificial intelligence, but designing experiences that, with artificial intelligence, enable them to continue learning, maturing, and building judgment and expertise.
We can shorten the time needed for the task. We can shorten the time spent on research. We can shorten the first draft.
But we must be very careful… not to shorten the journey of building expertise itself.