If AI Takes the Entry-Level Work, Where Do Your Future Leaders Learn?

If AI takes the entry-level work, your future leaders learn by supervising it instead of doing it: checking AI output, handling the exceptions, and sitting closer to customers and decisions earlier. That only happens if you design it on purpose. Automate the rote work without a plan and you quietly remove the training ground.

How will future leaders learn if AI takes entry-level work?

They learn by supervising the work instead of doing it by hand. Have them check what the AI produced, handle the exceptions, and sit closer to customers and decisions earlier. That only happens if you design it on purpose. Automate the rote work with no plan, and you quietly remove the training ground.

An operator I talked with recently raised a concern I keep thinking about. AI is killing entry-level and early-stage jobs that are more rote. If we're taking away the training grounds, how will we ever have senior leadership?

It's a fair question. And it's one most companies aren't asking yet.

What the rote work used to teach

My first sales job was at a semiconductor brokerage, making a lot of calls a day. It wasn't glamorous. But it taught me how buyers talk, how deals move and what the business actually depended on.

That's what rote work did for a lot of people. Data entry taught you what the data meant. Building the report taught you what leadership looked at. Answering the phones taught you what customers cared about. Nobody called it training. It was.

The tension: efficiency versus the pipeline of people

On one hand, automating rote work makes sense. It's slow, error-prone and nobody loves doing it. On the other hand, if you remove it entirely, you remove the place where people learned how the business works.

I land in the middle. Automate the work. Don't automate away the learning.

Move juniors up the work, not out of it

The approach that makes sense to me is to shift junior roles from doing the work to owning it:

• Review the output. Someone has to check what the AI produced. That's real learning.
• Handle the exceptions. The cases automation can't handle are where judgment is built.
• Explain the why. Ask juniors to explain why something was done a certain way.
• Improve the process. Let them suggest changes to how the automation works.

This keeps people close to the work, without making them do it by hand.

Design learning on purpose

When learning used to happen by accident, you didn't need a plan. Now you do. Think about what your future leaders need to understand, and build it in:

• Time with customers early, not after three years.
• Rotations through different functions.
• Sitting in on decisions, with someone explaining the reasoning.
• Real responsibilities with feedback, even small ones.

Don't stop hiring juniors

Some companies are cutting entry-level hiring altogether. It saves money this year. In five or ten years, they'll have nobody who grew up in the business. A smaller number of well-developed juniors working alongside AI is a better bet than none.

Signs you're removing the training ground

• Junior roles have shrunk to checking boxes in a tool, with no exposure to why.
• New hires rarely talk to a customer in their first year.
• Your senior people can't name who would replace them in five years.
• Hiring plans for the next year include no entry-level roles at all.

What to do this week

List the rote tasks your team has automated, or plans to, in the next year. Next to each, write one thing a junior person used to learn from doing it: how pricing works, what customers complain about, which data matters. Then decide where each lesson will be learned now. Maybe it's reviewing the AI's output weekly. Maybe it's sitting in on customer calls. If a lesson has no new home, it will quietly disappear, and you'll notice in a few years when nobody is ready to step up.

A worked example

Say a finance team automates invoice matching, a task juniors used to spend hours on. Done carelessly, the junior role becomes "click approve." Done well, the junior reviews every exception the automation flags, writes a short weekly note on the patterns they see, and joins the monthly close meeting to explain them. They're doing less typing and learning more about how the business actually runs.

Common mistakes

• Treating review work as busywork. Checking AI output is real learning if someone explains the why.
• Cutting every junior role at once. You can't hire ten years of experience later.

Let AI handle the busywork, keep people on the judgment

At Ants, this is how we think about AI agents. They should take the follow-ups, reminders and routine tasks — the work that never gets done — so your people can spend their time on the judgment, relationships and decisions that actually build careers.

Ready to put a colony on the work slipping through the cracks? Ants builds, runs and maintains the AI agents that handle follow-ups, renewals, invoices and deadlines, so your team can focus on the work that grows people and the business. Book a Colony Call.

Questions people ask