Research opportunity · Levelling or hollowing? Skill compression and the reproduction of expertise
What work do junior staff in Greek firms still learn by doing once AI arrives?
In short: has this been done?
Related work: Partly addressed. Related studies cover parts of this question; what they leave open is set out below. Only a first-pass search has been done, so it is not yet known whether a fuller search would find a study that does the whole design.
How far checked: First-pass screen only. A first-pass AI web search on 18 October 2026. No deeper check yet. Every search and check so far was done by AI; no person or domain expert has checked them yet.
The searches, the nearest work found and what the search could not reach
This page shows RO-LD04-08 version v1, whose record never changes. It is the current version; if the opportunity is revised, the newer version appears at its main page.
Question
In Greek accounting offices, law firms and software companies that use generative AI, which tasks that juniors used to learn on are now done with AI or by seniors, and how do juniors and their supervisors describe the feedback and practice juniors now get?
Gap type: Untested beyond the original settingUndergraduate thesisLow costRelated work: Partly addressed
Why it matters
LD-04 asks whether AI removes the work through which novices become experts. The evidence it cites is about hiring in the United States and Sweden, an ethnography of surgical training and a theoretical model; no study SocioLens found measures whether the tasks juniors learn on are shrinking, and the interview studies found are of software engineers in South Korea and elsewhere. A small interview study cannot say how common a change is, but it can describe whether and how it happens in occupations and a country the debate does not cover, and give a later survey better questions.
Evidence that motivates it
- In US résumé data covering 65 million workers, firms that began integrating generative AI saw junior employment fall by about 9% relative to senior employment after six quarters, compared with non-adopters, mainly because they hired fewer juniors, while junior promotion rates rose slightly. (LD04-SL-C15 v1.0)
How this study uses it: In US résumé data, firms that began integrating generative AI saw junior employment fall about 9% relative to senior employment, mainly through fewer junior hires. - A two-year comparative ethnography of robotic and open surgery, with interviews at 13 US teaching hospitals, found that approved routes for learning failed in robotic surgery; only a minority of trainees came to competence, through tolerated but unapproved practices, and these practices led to hyperspecialisation and a shrinking supply of experts relative to demand. (LD04-SL-C18 v1.0)
How this study uses it: When surgical robots took trainees out of hands-on work, approved routes for learning failed and only a minority of trainees reached competence, through tolerated but unapproved practice. - A theoretical model in which novices acquire tacit knowledge by working with experts shows that automating entry-level work can raise output yet lower long-run growth, even without cutting entry-level employment, by moving novices away from the most productive experts. (LD04-SL-C19 v1.0)
How this study uses it: A theoretical model shows that automating entry-level work can lower long-run growth even without cutting entry-level jobs, by moving novices away from the best experts. - In two preregistered lab experiments and a field study, LLM access had no overall effect on learning: students who used it to generate solutions covered more topics but understood each less, students who asked it for explanations understood more, and the gap between students with low and high prior knowledge widened. (LD04-SL-C13 v1.0)
How this study uses it: Students who used a language model to generate solutions understood less than those who asked it for explanations, so how juniors use the tool may matter as much as whether they use it.
Each claim is shown as it read when this version was published.
A possible design
Semi-structured interviews with about 12 to 16 people in two or three occupations (for example accounting, law and software), pairing where possible a junior with under three years in the job and a supervisor in the same firm. Ask about concrete recent tasks: who did each step, with which tool, who checked it and what feedback followed, and how the same task was done before generative AI. Analyse with thematic analysis and compare juniors' and supervisors' accounts. Report the findings as descriptions of what happens in these firms, not as estimates of how common it is.
Data needed
Access to 12 to 16 interviewees through the university careers office, professional associations or personal networks; consent and anonymisation under the university's ethics procedure; recording and transcription (free tools are enough).
Feasibility
Undergraduate thesis. Very low; interviews in person or by video call and free transcription software, with no specialist data. Suited to a final-year sociology thesis with training in qualitative methods. Firms may be reluctant to discuss AI use, so recruit individuals rather than firms, and ask for examples rather than opinions. It complements RO-LD04-03 and RO-LD04-04, which need larger designs.
Has this been done?
Partly done already: related work covers some of this question (first-pass screen of 18 October 2026; not yet checked in depth).
Interview studies already describe this mechanism among junior and senior software engineers in South Korea (14 interviews) and among 13 junior software developers whose country the abstract does not state. No interview study was found in Greece, or of junior accountants or lawyers, that pairs juniors with their supervisors and asks what happened to specific tasks juniors used to learn on. The search was web-based and partial.
How strong this evidence is
- Who searched and checked: no checker is recorded beyond the first-pass screen, which was part of the AI drafting (Claude (SocioLens Project Lead), AI-drafted from the LD-04 debate as an undergraduate-scale project, after the 17 October 2026 reader tests found no opportunity a final-year student could take on). Every check so far was done by AI; no person or domain expert has checked it yet.
- What was read: 2 of the 2 nearest works below were judged from an abstract, summary or search listing only; each is marked. Not every entry says how much of the work was read.
- Where the search could not reach: Screened by three web searches (two in English, one in Greek) and arXiv records on the date shown; Greek university repositories, the National Archive of PhD Theses and education and management databases were not searched. Absence of a hit is not proof that the work does not exist.
- Before building a project on it, run your own search in a scholarly database such as Scopus, Web of Science, Google Scholar or SSRN.
Nearest work found
- Yu and Moon, "Who Will Become the Next Senior? How Generative AI Erodes the Development Pathway in Software Engineering", arXiv:2607.17067 (2026; accepted to AIES 2026) Judged from an abstract, summary or search listing only
Abstract read. Fourteen interviews with junior candidates and senior software engineers in South Korea; the authors find that generative AI redirects entry-level work into senior engineers' AI-assisted workflows, so juniors lose the productive struggle that built expertise. Software only, one country. - Andriotte and Ribeiro, "Verification-Conditioned Use: A Qualitative Study on How Generative AI Reshapes Learning, Autonomy, and Market Entry for Junior Software Developers", arXiv:2607.24606 (2026) Judged from an abstract, summary or search listing only
Abstract read. Thirteen interns and junior developers interviewed by video call; whether they use AI depends mainly on whether they can check its output, and the authors describe shallow learning that weakens the judgement employers want. The participants' country is not confirmed, supervisors were not interviewed, and the study covers software only.
Search queries used
interviews junior professionals generative AI training apprenticeship tasks delegated qualitative study 2025Greece generative AI junior employees qualitative interviews workplace learningνέοι εργαζόμενοι τεχνητή νοημοσύνη εκπαίδευση στη δουλειά junior λογιστές δικηγόροι συνεντεύξεις έρευνα Ελλάδα
Cite this page
SocioLens project. (2026, October 18). What work do junior staff in Greek firms still learn by doing once AI arrives? [Research opportunity RO-LD04-08, version v1]. SocioLens. https://sociolens-web.pages.dev/opportunities/ro-ld04-08/v1/
@misc{sociolens_ro_ld04_08_v1,
author = {{SocioLens project}},
title = {What work do junior staff in Greek firms still learn by doing once AI arrives?},
year = {2026},
date = {2026-10-18},
howpublished = {\url{https://sociolens-web.pages.dev/opportunities/ro-ld04-08/v1/}},
note = {Research opportunity RO-LD04-08, version v1}
}The link points to this version, whose record never changes. For a finding, cite the original study as well: SocioLens summarizes research and does not replace it.
Provenance
RO-LD04-08 v1. Drafted by Claude (SocioLens Project Lead), AI-drafted from the LD-04 debate as an undergraduate-scale project, after the 17 October 2026 reader tests found no opportunity a final-year student could take on on 2026-10-18. Not yet reviewed. Not yet reviewed by a domain expert. This is a research hypothesis, not a finding.