# AI 'Employees' Might Disrupt Work in Unexpected Ways > [!question] Lingering Questions > - The full NYT article is paywalled; only a ~297-word preview was retrievable. Both source papers referenced (Traberg on research homogenization; Wiles on AI-employee framing) need to be located and read directly before anything here is promoted further. > - How does Traberg's paper actually operationalize "homogenization" — lexical variance, topic clustering, something else? > [!tip] Article Ideas > - none held — too thin a source to develop independently; useful only as a supporting citation once the full article or underlying papers are in hand. > [!link] Cross Reference > - [[3. Resources/Permanent Notes/AI's Failure Mode Is Careless Adoption Not the Technology]] — the vault's existing Permanent note for this exact source. > - [[3. Resources/Literature Notes/Agentrification (Scott Logic)]] — makes a structurally similar claim (individual-level AI use looks fine; collective-level effects are what create the real cost) from a different angle. > [!quote] Vocabulary > - none new — preview is too short to introduce any coined terms. ## Bib capsule New York Times business-section reporting by Noam Scheiber, published 2026-06-29. **Paywall limit**: this note covers only the accessible preview (~297 words / ~1,690 characters); the full article continues behind the paywall and could not be retrieved. No findings beyond the preview are inferred. ## Author's thesis (what it's about / what's said in detail) The preview reports two research strands on organizations absorbing AI into knowledge work: one on research homogenization (heavy AI use across all stages of scholarly work makes individual output look fine but collective output "sound alike"), and one on managing "AI employees" — roughly a quarter of surveyed organizations now list AI agents as peer entries on org charts. The preview's closing note, from a Boston University researcher, locates AI's shortcomings in careless human adoption practice rather than in the technology itself — but the preview cuts off before reporting the labeling experiment's actual results. ## Key arguments (≤3) 1. **Individually invisible, collectively visible** — AI-assisted research looks fine scholar-by-scholar but produces measurable convergence ("everyone sounding alike") in aggregate. 2. **Organizations are starting to treat AI agents as org-chart peers**, not tools — which changes supervision norms, though the preview doesn't show how. 3. **The locus of AI harm is adoption practice, not model capability** — this is the preview's most load-bearing claim, and already fully captured in the vault's existing Permanent note. ## Claim atoms ### AI at every research stage homogenizes scholarship collectively > [!quote] ID:rw_preview > But at the collective level, it looks pretty similar. Everyone is sounding alike. Where/why: Traberg (Copenhagen Business School), on AI used across all stages of a research pipeline — invisible to any one scholar, visible only in aggregate. ### Organizations label AI agents as org-chart peers > [!quote] ID:rw_preview > It's technically an equivalent peer on your team. Where/why: an anonymized manager describing an agent named "Scout"; the preview reports roughly a quarter of surveyed organizations list AI agents on org charts. ### AI's failures arise from careless human adoption > [!quote] ID:rw_preview > the shortcomings weren't necessarily intrinsic to the technology, but arose when humans adopted it with little attention to what could go wrong. Where/why: Dr. Wiles (Boston University), reframing AI-governance as a human-process problem rather than a model-capability one. ## Permanent candidates - **[[3. Resources/Permanent Notes/AI's Failure Mode Is Careless Adoption Not the Technology]]** — already exists in the vault (created 2026-06-30, same origin doc). Covers claim atom H3, including the labeling-experiment mechanism (managers caught fewer planted errors when told an "AI employee" made the work) in its Scope & conditions. Not recreated here. ## Processing checklist - [x] Checked for an existing Permanent note before drafting — found one covering the load-bearing claim; linked instead of duplicating. - [ ] Get the full NYT article and both underlying papers (Traberg; Wiles) once paywall-accessible — this note stays `seedling` until then. - [x] Connected into [[3. Resources/MOCs/AI & the Future of Knowledge Work MOC]]. ## Links - [[3. Resources/Readwise/Articles/A.I. ‘Employees’ Might Disrupt Work in Unexpected Ways|A.I. 'Employees' Might Disrupt Work in Unexpected Ways]] — full saved source document (preview only) - [[3. Resources/MOCs/AI & the Future of Knowledge Work MOC]] — org-chart-peer framing and adoption-practice-as-locus-of-harm both belong to the vault's broader AI-at-work thread