KI tar ikke jobbene, men inngangsporten

Lasse Maugesten · article · 2026-04-17 source ↗ Reader 🖥️

AI Labor Disruption Systemic Apprenticeship & Skill Transmission

AI is not reducing overall employment, but it is closing the entry-level pathway into careers by automating the routine, book-knowledge tasks junior employees used to learn from — so unless this is addressed, today's shrinking junior intake will become tomorrow's shortage of experienced workers, in Norway as elsewhere.

practical · argumentative — A short Norwegian opinion/analysis piece (E24, written by Lasse Maugesten of communications agency TRY) arguing that AI is not destroying jobs overall but is specifically closing the entry-level pathway into careers — illustrated by one new graduate's eight-month job search, Randstad hiring data, a Stanford labor study, Norges Bank data, and the author's own agency's experience.

Structure

Part	Role
Opening case: Ida, a 25-year-old NHH graduate, unemployed after six interviews and eight months of searching	narrative hook / evidence
Randstad data: entry-level roles (0-2 years experience) down 29 percentage points since January 2024	evidence
Stanford labor economists: 13% employment decline for 22-25-year-olds in AI-exposed occupations since ChatGPT's 2022 launch, vs. 5-8% increase for older workers in the same occupations	evidence
Mechanism: AI replaces codified/book knowledge, not tacit knowledge gained only through years on the job; juniors have the most of the former and least of the latter (NHH's Bram Timmermans)	explanation
Norwegian evidence: 55% of Norwegian businesses now use AI (up from 24% in 2023); Norges Bank's December 2025 report links increased AI use to fewer jobs, especially in data-analysis- and writing-heavy roles	evidence (localizes the claim to Norway)
Author's own agency (TRY): a senior developer now holds two complex projects in mind at once with AI help; a strategist delivers in one day work that used to take a three-person team two weeks — but only those who already know what they're looking for gain this leverage	evidence / mechanism detail
Closing argument: today's juniors are tomorrow's seniors; closing the entry gate now creates a future shortage of the experienced people everyone agrees are needed	conclusion

Key terms

inngangsporten (the entry gate/pathway)
The set of routine tasks and junior roles — data collection, analysis, quality assurance, first drafts — through which new graduates traditionally built competence; the author's central claim is that AI is closing this specific gate even while overall employment looks unaffected.
boklig kunnskap vs. taus kunnskap (book knowledge vs. tacit knowledge)
Book/codified knowledge is what AI replaces — what you learn in school; tacit knowledge is what you can only acquire after years on the job, and AI does not replace it. Juniors have disproportionately more of the former and less of the latter, which is why they are disproportionately exposed.

Propositions

explicitEntry-level roles requiring zero to two years of experience have fallen 29 percentage points globally since January 2024, based on Randstad's analysis of 126 million job postings.conf: high
explicitAmong 22-25-year-olds in AI-exposed occupations (software development, customer service, analysis), employment has fallen 13% since ChatGPT's late-2022 launch, while employment for older workers in the same occupations rose 5-8% over the same period.conf: high
supportsThe loudest warning about AI and jobs yet · p2 — Independent corroboration (Stanford labor-economics data on 22-25-year-olds in AI-exposed occupations) of the same early-career-specific displacement pattern reported in the US jobs-crisis piece.
explicitAI replaces codified/book knowledge but not tacit knowledge accumulated only through years on the job; since junior employees disproportionately hold the former and lack the latter, they are disproportionately exposed to AI displacement.conf: high
explicitThe routine tasks now being automated away (data collection, analysis, quality assurance, first drafts) were not just junior busywork — they were the learning situations through which new graduates built the competence that makes them valuable a decade later; as AI and AI-equipped senior staff absorb these tasks, the path from education to competence lengthens.conf: high
explicitIn Norway specifically, AI use among businesses rose from 24% in 2023 to 55% currently, and Norges Bank's December 2025 report finds this increased use is already reducing the number of jobs, particularly in data-analysis- and writing-heavy roles — even though this effect is not yet visible in Norway's aggregate unemployment statistics.conf: high
explicitThe AI-driven capacity gain is unevenly distributed within organizations: those who already know what they are looking for get multiplied capacity from AI tools, while those who lack the experience also lack the foundation to use the tools effectively — meaning AI amplifies existing experience rather than substituting for its absence.conf: high
explicitBecause today's juniors are tomorrow's seniors, closing the entry gate now will, within about a decade, produce a shortage of experienced workers that all sides agree the economy needs — turning a present cost-saving into a future structural deficit.conf: high

Arguments

The entry-level job market is closing specifically because of AI substitution for codified knowledge, not primarily because of a general economic downturn (konjunktur). strength: high

Explicit premises:

If this were a general economic cycle, the decline would not be concentrated in roles with 0-2 years of experience specifically.
Randstad's global data shows a 29-point decline concentrated exactly in entry-level roles.
Stanford's wage data shows the decline concentrated in AI-exposed occupations and in the youngest age cohort within them, while older workers in the same occupations gained.

Implicit premises:

A pattern this specific (youngest workers, in AI-exposed occupations, falling while older workers in the same occupations rise) is better explained by AI substitution than by any plausible alternative (e.g., generational hiring preferences, credential inflation, or cyclical effects that happen to concentrate on juniors).

Inferences:

If the mechanism is AI substitution for codified knowledge rather than a cycle, the effect should not reverse automatically when the macroeconomy improves — it requires a deliberate organizational response (e.g., redesigning how juniors learn) rather than just waiting it out.

Narratives

Ida Fronth-Andersen, a 25-year-old business graduate from NHH who studied in China, has been job-hunting since August 2025 (eight months as of publication), with six interviews and three times finishing second. She has 'done everything right' — right education, right school, engagement, will — yet remains unemployed.
→ Personalizes the statistical claim (29-point drop in entry-level roles, rising graduate unemployment) and sets up the article's closing point: the system meant to admit people like Ida is closing, independent of individual merit or effort.
At TRY (the author's own agency), a senior developer now holds two complex projects in mind simultaneously because AI handles what used to require full solo concentration; a strategist analyzes three years of communications data and delivers a recommendation in one day — work that previously took a three-person team two weeks.
→ Shows the mechanism concretely from the inside: AI multiplies the capacity of people who already have the experience to direct it, which is the flip side of why junior employees (who lack that experience) are being squeezed out rather than uplifted.

Questions addressed

✅ Is the drop in entry-level hiring just a normal economic downturn (konjunktur), or something structurally different?
The author argues it is structurally different — specifically AI-driven — based on the Randstad and Stanford data showing the decline concentrated in entry-level/youngest/AI-exposed segments rather than spread evenly.
✅ Why would AI disproportionately affect junior employees rather than affecting all experience levels equally?
Because AI replaces codified/book knowledge (which juniors have most of) rather than tacit knowledge (which juniors have least of and which only comes from years on the job).
⭕ What should organizations or policymakers actually do to keep the entry gate open?
The article poses the question explicitly in its closing line ('Spørsmålet er hvor de skal komme fra når ingen slipper inn') but does not propose an answer.

Gaps

incompleteThe article diagnoses the problem sharply (entry gate closing, future senior shortage) but names no candidate response — no organizational redesign of junior work, no policy lever, no example of a company or sector doing this differently. It ends on the open question rather than any direction toward an answer.
Read next: Documents elsewhere in this corpus that propose concrete organizational responses to the same mechanism (e.g. protected mentor time, deliberately structured AI-free learning tasks, or redesigned internship programs) would be the natural next read.
incompleteThe TRY examples (senior developer, strategist) are offered as illustrations of AI amplifying existing expertise, but the article doesn't explore whether an organization could deliberately engineer a substitute path to tacit knowledge once the traditional one (junior grunt work) is gone — e.g. simulation, shadowing, or AI-assisted mentorship designed specifically to transmit judgment rather than just output.
Read next: Look for proposals of a deliberately engineered 'synthetic apprenticeship' or structured AI-assisted mentorship model.

Closes gaps elsewhere

The loudest warning about AI and jobs yet · g3 — Provides exactly the Nordic/European data point that gap asked for — and the answer is the opposite of the gap's speculation: Norway shows the same junior-pipeline closure (rising new-graduate unemployment, Randstad's entry-role decline, and Norges Bank's December 2025 finding that AI use is already reducing jobs), so stronger European safety nets have not so far provided the cushion the gap hypothesized.