"The bottom rung is gone" "Your AI doesn't need onboarding" "Who replaces the replacements?"

Restarting Article #11.

Subject line: The bottom rung is gone Subtitle: Entry-level roles are disappearing faster than anyone is replacing them. Every organisation with a talent pipeline should be paying attention.

Artificial Idea | AI careers · practical prompts · no hype Monday, September 8, 2025 · Issue #11 · Jobs

The pipeline problem

This is not just a problem for young professionals. It is a structural crisis for every organisation that has ever promoted someone from within.

In the first half of 2025, LinkedIn's Global Talent Trends report recorded a 24% year-on-year decline in entry-level job postings across financial services, technology, legal, and marketing sectors in G20 economies. In the same period, postings for senior and specialist roles in those same sectors fell by just 4%. The labour market is not contracting uniformly. It is contracting from the bottom, and the consequences of that pattern extend well beyond the young professionals who cannot find their first job.

To understand why, you need to understand what entry-level roles actually do inside organisations. The obvious answer is that they produce work. Junior analysts run reports. Junior copywriters produce first drafts. Junior developers write boilerplate code. That output has value, and AI is increasingly capable of producing a version of it faster and more cheaply.

But entry-level roles produce something else that AI cannot replicate, something that does not appear on any balance sheet and is only noticed when it stops being produced. They develop the senior professionals of the next decade.

How talent actually gets built

The McKinsey Global Institute's 2025 workforce development report makes a finding that should be uncomfortable reading for any senior leader who has benefited from a traditional career progression. It found that 73% of the skills that distinguish high-performing senior professionals were acquired not through formal training but through the accumulation of low-stakes, high-volume work in early career roles. The junior analyst who ran a thousand reports learned, through that process, how to spot anomalies, ask better questions, and develop instincts about what data actually means. The junior lawyer who reviewed thousands of contracts developed a pattern recognition capability that no law school curriculum and no AI tool currently produces.

That learning process requires repetition. It requires making small mistakes in low-consequence situations. It requires the gradual accumulation of domain intuition that is, by its nature, experiential rather than instructional. Entry-level roles are the infrastructure through which that accumulation happens. When those roles disappear, the infrastructure disappears with them.

The result, which will not be fully visible for another five to seven years but is already beginning to manifest in early indicators, is a hollowing of the professional pipeline. Organisations are eliminating the roles that produce their future senior talent while simultaneously competing for the senior talent that those roles would have developed. The contradiction is structural and it is largely unacknowledged in the boardroom conversations driving the automation decisions.

The early indicators are already visible

A 2025 survey of 600 hiring managers across the United States and United Kingdom conducted by the Society for Human Resource Management found that 67% reported increasing difficulty finding mid-level candidates with the foundational skills their roles require, despite those roles offering competitive compensation. The most commonly cited gap was not technical. It was the ability to exercise independent judgment on moderately complex problems without requiring senior oversight.

That capability gap does not emerge from nowhere. It is the predictable downstream consequence of a talent pipeline that has been progressively compressed at the entry point over the preceding three to five years. The organisations reporting the most acute mid-level talent shortages in 2025 are disproportionately concentrated in the sectors that automated entry-level roles most aggressively between 2021 and 2023.

Correlation is not causation, and the full causal picture is more complex than a single variable. But the pattern is consistent enough, and the mechanism clear enough, that dismissing the relationship requires more evidence than currently exists on the other side of the argument.

What this means for young professionals specifically

For individuals early in their careers or approaching the start of one, the disappearance of traditional entry-level roles creates a specific and serious challenge. The conventional pathway, where a degree leads to a junior role that leads to a mid-level role through demonstrated competence, is narrowing in precisely the sectors that have historically offered the most structured development, the clearest progression, and the strongest compensation.

The practical response to this is not to avoid those sectors. It is to understand that the entry points have shifted. Organisations that are not hiring junior analysts are hiring people who can demonstrate analytical capability without needing to develop it on the job. The bar at entry has risen because the volume of entry-level positions has fallen, and the candidates who clear it are those who have found alternative routes to building the foundational skills that entry-level roles previously developed.

Internships, freelance projects, self-directed AI-augmented work, open-source contributions, and visible independent projects are increasingly functioning as the new entry-level infrastructure. Not because organisations designed it that way, but because the compression of formal entry-level hiring has created a vacuum that motivated candidates are filling through other means.

A 2025 analysis by the National Bureau of Economic Research found that candidates who could demonstrate applied experience through independent projects were 3.2 times more likely to receive first-round interview invitations for competitive junior roles than equally qualified candidates with equivalent academic credentials but no applied portfolio. The credential still matters. The portfolio is increasingly what gets you in the room.

What this means for organisations

For leaders responsible for talent strategy, the entry-level compression creates a problem that is easy to defer and expensive to ignore. The organisations that automated their way through junior headcount in 2022 and 2023 are beginning to discover that the mid-level talent they need in 2025 and 2026 was not being developed anywhere. It was being developed in the junior roles they eliminated.

The organisations navigating this most effectively are those that have redesigned rather than eliminated entry-level roles, keeping humans in the pipeline but reorienting their function from output production to capability development, with AI handling the production work and junior professionals handling the judgment, quality control, and learning that AI cannot replicate. This is not a cost-neutral approach. It requires accepting that some junior roles will produce less immediate output than an AI alternative would. The organisations making that investment are treating it as talent infrastructure spending, not operational cost. The ones that are not are borrowing against a future they have not fully accounted for.

The action

If you manage people: audit whether your team still has a genuine entry point. Not an internship that produces coffee and slide formatting. A role through which a motivated person with foundational skills could develop into someone you would fight to keep in three years. If that role does not exist, you are outsourcing your future talent pipeline to your competitors who still have one.

If you are early in your career: the portfolio is not optional anymore. The credential opens the conversation. The demonstrated applied work closes it.

Thursday we give you the prompt framework that early-career professionals are using to build applied AI portfolios quickly, credibly, and without needing access to a corporate environment to do it. The approach is specific, the results are visible, and the candidates using it are standing out in a market where most of their peers are still relying on credentials alone.

— The Artificial Idea team