If AI Automates the Apprenticeship, Who Becomes Senior?

Picture this: a customer service team at a company called Northstar handles around 20,000 questions every month. Most of those questions are not particularly exciting. “Where is my order?” “Can I change my address?” “Why have I been charged twice?” You get the idea. Northstar introduces AI, and six months later, the system can handle roughly 30% of the work. What happens after AI automates the first level of tasks?

What should the company do next?

Well, the short answer to this (and to most questions involving people, technology, and a suspiciously confident spreadsheet) is: it depends (what a radical statement for a consultant).

Northstar could remove 30% of the team. Nice, clean, and wonderfully easy to explain in the next board meeting. Or it could use the newly available capacity to respond faster, contact customers before problems escalate, and let people handle the difficult cases that were previously rushed. Same technology. Same productivity gain. Yet a completely different decision.

AI does not decide what happens to the job

We often talk about AI as if it enters the building, finds an empty desk, and starts handing out redundancy letters. It does not. Leaders make those decisions. “AI will take my job.” Although I understand the sentiment, it ultimately is a decision that is deliberately made. And I am fully aware that the vast majority of organizations are not handing out occupational therapy. They should obviously be profitable.

Jobs are collections of tasks. AI might draft the report, summarize the call, reconcile the spreadsheet, or provide a first answer. That does not automatically mean it can take over the judgment, accountability, and context wrapped around those tasks.

The International Labour Organization estimates that one in four workers worldwide has a job with some exposure to generative AI. Scary headline? Sure. But it concludes that most of those jobs are more likely to change than disappear because human input is still required.

The World Economic Forum predicts plenty of disruption by 2030: 170 million roles created and 92 million displaced across several major trends. Looking specifically at AI and information-processing technologies, it expects 11 million roles to be created and 9 million to be displaced.

So, more jobs then? Problem solved, and everyone can go home ea... NOPE!

Global job numbers will not help the person whose role disappears on Tuesday. They also tell us very little about whether companies are redesigning work well or simply cutting costs quickly (cutting costs obviously sounds a lot sexier in the ears of many investors and board members).

What if AI Automates the Apprenticeship

Another issue hides beneath the efficiency discussion. Junior work is often repetitive. It is also about how people learn. Analysts develop judgment by checking numbers. Lawyers spot patterns by reviewing documents. Designers build taste through the versions that never make it in front of a customer.

If AI removes all of that work, how does someone become senior? I, for sure, would be nowhere without my junior years, and I would be dumb as hell; I can assure you that.

This does not mean we should preserve pointless tasks for sentimental reasons. It does mean that removing a task and removing its learning value are two different decisions. If you automate the apprenticeship, you need to design a new one.

Five questions worth putting on the wall

Before announcing a headcount target, make the work transparent. Start with these questions:

None of these questions will provide value by themselves. The value lies in the conversation that follows and the decisions leaders make because of it. That is where the future of work is being shaped.

When AI automates junior tasks, it will remove some jobs and create others. But inside your organization, “the technology made us do it” is not a strategy. It is an excuse. Remember the statement in the IBM training manual: “A computer can never be held accountable; therefore, a computer must never make a management decision.”

The future of work is not happening to business leaders. They are designing it. They might as well do that transparently.