The Line I Almost Missed
The Intern, the Agent, and What Stayed Human
A Teams call with a young colleague left me staring at a task list and an uncomfortable thought: I was beginning to manage people the way I manage agents.
Teams call with a young team member ended. The window closed. The room went quiet. The task list stayed on my screen.
We had just gone through a few background tasks. The kind of work that always exists in a team, even when nobody feels like touching it.
Normal management work.
Put something in the background. Ask for a first pass. Agree to come back with feedback. End the call.
Then the thought arrived: I am starting to work with people like agents.
That was the part that unsettled me.
I understand the appeal of an agent.
It does not forget. It does not lose a task. It does not break the basic contract of its role.
For recurring operational work, checks, drafts, reminders, follow-through, it is reliable exactly where people sometimes are not. Not because the work is hard. Because attention runs thin. Motivation drops. Things slip.
I have built up a certain fatigue around that pattern.
Not toward specific people. Toward repetition.
Explaining the same thing again. Seeing basic things left undone. Asking for a change, getting agreement, then watching nothing move.
And this is where the thought became uncomfortable. I could hand almost any of those tasks to an agent now. Not someday. Now.
That is useful. But it also changes something in me.
If I am not careful, I stop seeing the difference between a person and a reliable background executor. I start valuing both through the same lens: did the task move, did the check happen, did the draft appear.
That is when the line almost disappeared.
What brought it back was not a perfect deliverable.
Later, I could not even clearly remember the exact idea he had suggested. What stayed with me was something else. The tone in his voice. The speed with which he picked up the topic. The feeling that it actually mattered to him.
That difference is easy to miss if all you look at is output.
An agent can return a draft.
A person can return involvement.
That is not a small difference.
That is the difference between execution and presence.
Care. Surprise. Energy. Interest. Responsibility. A bit of drive in the work. Those things do not always show up as a better first draft. But they change what a team feels like from the inside.
That was the real fear.
Not that agents will replace people.
That I might slowly start managing people as if those things did not matter.
I do want agents in the team.
I want them for the operational load that drains attention and patience. I want them for recurring checks, background drafts, and the work that should happen reliably without emotional overhead.
But I also want live people in the team.
Not as decoration around an efficient system. Not as a softer interface around automation. As the source of something the system does not generate on its own.
So the question for a manager is changing.
Not: can an agent do this task?
Often, yes.
The harder question is: what do I still want from a person once reliability can be delegated?
If I do not answer that honestly, I will build the wrong kind of team, even if it looks efficient from the outside.
The line did not disappear.
For a moment, I almost stopped seeing it.