Skip to content
Das GenZ-Dilemma
ArticlesAI translation· Das GenZ-Dilemma · Teil 3· 26 June 2026

The GenZ Dilemma, Part 3: The “Seniorisation” of Juniors

by 6 min read

Regular readers saw this coming, and now the empirical evidence is hardening: the job market for career starters is mutating. Recent analyses from Stanford and Harvard, along with the hard US labour market data from spring 2026, paint a clear picture. The number of entry-level jobs has dropped sharply (in some US sectors by up to 35%), while underemployment among fresh graduates has hit a worrying high.

But the real crisis isn't purely quantitative. This isn't about job cuts, because hiring data for senior roles is stable or even rising. What we're seeing is a fundamental structural shift in the career ladder: the “seniorisation” of the junior role.

The Analysis: The Paradox of the Perfect Entry-Level Hire

So what's actually going on? Senior developers, experienced lawyers and strategy consultants are using generative AI to do the work they used to need one or two juniors for. The first code iteration, summarising market data, the first contract draft, the classic “grunt work”: all of it is now handled by models that are faster, cheaper and (at this level) relatively error-free. To make the competition clearer: an AI makes mistakes, but it makes them faster than a junior. And when you steer it to fix those mistakes, it's faster at that too. The effort of this “trial-and-error” routine is lower with AI agents than with juniors, so the optimisation imperative wins out quickly.

Today's junior is therefore stuck in a perfect Catch-22: the routine tasks through which they used to build Tacit Knowledge (experiential knowledge), intuition and judgement by endless repetition are, de facto, being dismantled.

Instead, companies now want the unicorn: a junior employee on a junior salary who, from day one, operates with the strategic foresight and contextual judgement of a mid-level or senior person, orchestrating and validating the AI's output. The “doer” is supposed to become a “manager of AI agents” overnight.

Which leads us straight into what experts now call the “Talent Doom Cycle”: if we saw off the bottom rung of the career ladder, where exactly are the experienced seniors supposed to come from in ten years?

A Real-World Example: The Reluctant Code Reviewer

Picture a young software developer in the mid-2010s. She started out fixing small bugs, documenting legacy code or writing simple CRUD routines. Through hundreds of hours of “doing it herself”, she developed a feel for clean architecture and potential pitfalls.

Our junior developer in 2026 starts differently. The AI serves her a fully generated solution. Her task: “Find the subtle security holes in this 2,000-line AI construct you didn't write yourself, and adapt it to our undocumented company architecture.”

That's a classic senior task (system design & code audit). Yet today it's expected of career starters who have completely skipped the painful but instructive process of writing their own code from scratch.

The Solution: AI as the Architect of New Curricula

If AI is the problem because it eliminates the experience loops on the job, then AI itself has to become the solution, and it has to happen before young people enter the workforce. We need to flip teaching and curricula at universities and training companies accordingly.

Here are three concrete scenarios for how new teaching concepts and AI support could resolve the “seniorisation” dilemma:

Scenario 1: The Shift from “Creation” to “Evaluation” (Flipped Output)

As long as we teach students how to write an essay, a financial plan or a script from scratch (and punish them for using AI to do it), we're preparing them for a world that ended in 2025.
The new way: Curricula should start by default with an AI-generated draft. This draft contains, deliberately configured that way by the instructor, subtle logic errors, bias or gaps in context. From then on, the students' main task and the basis for their grade is the critical audit. They have to argue like a senior manager why the AI solution fails in a specific business context and what the architectural improvement looks like.

Scenario 2: AI as a Socratic Sparring Partner, Not an Oracle

We have to stop seeing AI in education as merely a “smarter Google”. We can build dedicated agents that master the Socratic dialogue.
The new way: Instead of a business student simply crunching the numbers on a market analysis, he enters a chat dialogue with an AI mentor. The AI doesn't hand out answers; it asks pointed, provocative questions about the student's assumptions: “Your pricing model is based on X, but what happens if the supply chains in Taiwan collapse? Adjust your strategy.” Passive knowledge intake turns into a simulation of real, stress-tested project experience at breakneck speed.

Scenario 3: Synthetic “Grunt Work” and Apprenticeship in Fast-Forward

If companies have dismantled the once-protected workshop where mistakes were allowed, then universities and training companies have to simulate it.
The new way: AI-driven “business simulators”. A medical student or aspiring consultant interacts with hundreds of simulated AI patients or AI clients who deliberately give vague requirements, change their minds or deliver inconsistent data. Immersive role-play like this makes it possible to build the Tacit Knowledge that used to take years much faster, through compressed, intensive iterations of experience.

From Practice: A Pilot Project in Tunisia

We're currently testing whether these three scenarios hold up in practice in a first pilot project at the German Business School in Tunisia. There, we're trying to actively build this “seniorisation” into the curricula. Instead of dogmatically regulating the use of AI, we methodically put students in the role of evaluators. Their new task: orchestrate AI-generated concepts and audit them in realistic scenarios for errors, feasibility and business value.

We're still at the beginning, but the first interim evaluation makes us very optimistic. You can see systems thinking and critical thinking developing; the students are noticeably starting to drop passive “box-ticking” and train strategic questioning instead. Of course, the methodology and the exam formats still need work. But the trend so far points in a clear direction: methodically building “senior skills” in juniors is doable, provided we create the didactic framework for it and keep iterating on it.

Conclusion: We Have to Shape Tomorrow's Generations

Socially and economically, we can't afford to release Gen Z (and soon Alpha) unprotected into a job market that sees their traditional role as redundant while simultaneously demanding leadership poise.

Companies that look only at cost optimisation today and slash junior positions will have a massive problem at the top in five years. But those that realise now that training programmes must be turned into “structured AI apprenticeship programmes”, focused on learning judgement rather than mere task execution, will build the most resilient and innovative organisations of this decade.

AI has changed the rules of the career ladder.
We have to adapt.


Live long and prosper 😉🖖


(Subliminal)Soundtrack 😉

▶Listen