In an experiment where participants studied a technical domain with or without an LLM, gains were highly uneven and were predicted not by grades or prior knowledge but by skill at eliciting, filtering and verifying AI output; participants low on that skill saw limited or negative returns, and a concept-map scaffold reduced the spread.

SocioLens project,

Claim LD04-SL-C06 (v1.0)

Research opportunities this claim motivates

What this claim leaves open for research. Each is a hypothesis about a gap, not a finding.

What AI set-ups do employers actually give trainees and junior staff?A concept-map scaffold reduced the spread of gains from LLM-assisted study, an example of a design an employer could adopt.

Provenance and status

Claim type: Estimate
Publication status: Eligible for publication, with qualifiers: preprint
Verification state: Checked against the abstract of the primary source only. Who checked this claim is not recorded for it; claim checks so far were AI-led, and no person has checked claims against their sources.
Does not license: Workplace effects; a validated measure of AI interaction competence
Limitations: Preprint; student stand-ins; checked against the arXiv abstract only, so sample and magnitudes are not reported here
Source (as cited): Idan & Anand, “Generative AI and the Productivity Divide: Human-AI Complementarities in Education,” arXiv:2605.18143, May 2026
Source type: Preprint (recorded as “Preprint”)
Institutional position: Independent academic
Maturity of the evidence: Emerging evidence
Recorded date: 14 October 2026
Currentness / correction: This is the current version. No correction or concern is recorded on its sources.

Sources this version cites

Idan & Anand, “Generative AI and the Productivity Divide: Human-AI Complementarities in Education,” arXiv:2605.18143, May 2026 (Preprint)
Read the original

Version

This is the current version, v1.0. Earlier versions stay available, and their records never change.

Versions: v1.0 · current

Raw record: every recorded field, with its coded values

All 25 canonical Claim fields, verbatim, with the coded values derived from them. The plain-language labels above are read from these.

Claim type (coded): ESTIMATE (as printed: “ESTIMATE”)
Publication status (coded): ELIGIBLE (PREPRINT) (as printed: “ELIGIBLE WITH PREPRINT QUALIFIER”)
Verification state (coded): CONTENT_VERIFIED_PRIMARY (ABSTRACT_ONLY)
Source type (coded): PREPRINT
Maturity (coded): EMERGING

#	Canonical field	Canonical value (verbatim)	Normalized (Claim Value Vocabulary v1.0.2)
1	CLAIM ID	LD04-SL-C06	—
2	PUBLIC-FACING CLAIM	In an experiment where participants studied a technical domain with or without an LLM, gains were highly uneven and were predicted not by grades or prior knowledge but by skill at eliciting, filtering and verifying AI output; participants low on that skill saw limited or negative returns, and a concept-map scaffold reduced the spread.	—
3	CLAIM TYPE	ESTIMATE	base_types: ESTIMATE
4	RESEARCH STRAND	A — Performance while the tool is in use	—
5	MATURITY	EMERGING	states: EMERGING
6	SOURCE	Idan & Anand, “Generative AI and the Productivity Divide: Human-AI Complementarities in Education,” arXiv:2605.18143, May 2026	source_ids: SRC-SL-IDAN-ANAND-ARXIV-2605-18143
7	SOURCE TYPE	Preprint	classes: PREPRINT
8	INSTITUTIONAL POSITION	Independent academic	—
9	SCOPE	Self-study of a technical domain by participants standing in for early-career knowledge workers	—
10	DESIGN	Randomized experiment with a scaffolding intervention	—
11	SAMPLE / DATASET	Not checked	—
12	MODEL / TOOL / VERSION	LLM assistance (version not checked)	—
13	COMPARATOR	Traditional study resources	—
14	MAIN RESULT	Mean performance rises; gains highly uneven; “AI Interaction Competence” predicts gains, GPA and prior knowledge do not; low-competence participants see limited or negative returns; concept-map scaffolding reduces variance	—
15	INFERENTIAL TARGET	What predicts who benefits from LLM-assisted study	—
16	DOES NOT LICENSE	Workplace effects; a validated measure of AI interaction competence	—
17	VERIFICATION STATE	CONTENT VERIFIED ABSTRACT	state: CONTENT_VERIFIED_PRIMARY · qualifiers: ABSTRACT_ONLY · verified_by: UNKNOWN
18	LIMITATIONS	Preprint; student stand-ins; checked against the arXiv abstract only, so sample and magnitudes are not reported here	—
19	CONTRADICTORY / SUPPORTING SOURCES	Consistent with widening by interaction skill; implies that a gap can be narrowed by design	—
20	GEOGRAPHY / LANGUAGE	Not checked / English	—
21	DISTRIBUTIONAL LEDGER	Gains follow a new skill (using AI well) rather than prior attainment	—
22	R01 / FIELD-MAP LINKS	LD-04, D2	—
23	WHAT WOULD CHANGE THIS CLAIM?	Full-text reading or replication that changes the predictor result	—
24	PUBLICATION STATUS	ELIGIBLE WITH PREPRINT QUALIFIER	state: ELIGIBLE · qualifiers: PREPRINT
25	RECORDED DATE	2026-10-14	—

Field 22 references resolved against Field Map v0.2: LD-04 (LIVING_DEBATE); D2 (DOMAIN)

The sources as recorded, with their record IDs and coded source types:

Pinned Sources (field 6, version-specific)

“Idan & Anand, “Generative AI and the Productivity Divide: Human-AI Complementarities in Education,” arXiv:2605.18143, May 2026” → Idan & Anand (2026), “Generative AI and the Productivity…” (SRC-SL-IDAN-ANAND-ARXIV-2605-18143--v1.0) — Idan & Anand, “Generative AI and the Productivity Divide: Human-AI Complementarities in Education,” arXiv:2605.18143, May 2026; class PREPRINT (explicit public provenance pin)

Cited by Living Debate versions

Cite this page

SocioLens project. (2026, October 14). In an experiment where participants studied a technical domain with or without an LLM, gains were highly uneven and were predicted not by grades or prior knowledge but by skill at eliciting, filtering and verifying AI output; participants low on that skill saw limited or negative returns, and a concept-map scaffold reduced the spread [Claim LD04-SL-C06, version v1.0]. SocioLens. https://sociolens-web.pages.dev/claims/ld04-sl-c06/v1.0/

@misc{sociolens_ld04_sl_c06_v1_0,
  author = {{SocioLens project}},
  title = {In an experiment where participants studied a technical domain with or without an LLM, gains were highly uneven and were predicted not by grades or prior knowledge but by skill at eliciting, filtering and verifying AI output; participants low on that skill saw limited or negative returns, and a concept-map scaffold reduced the spread.},
  year = {2026},
  date = {2026-10-14},
  howpublished = {\url{https://sociolens-web.pages.dev/claims/ld04-sl-c06/v1.0/}},
  note = {Claim LD04-SL-C06, version v1.0}
}

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.

Research-of-record provenance

Materialized by sociolens-materializer 0.6.0 at 2026-09-28T20:00:00.000Z; normalization contract SocioLens — Claim Value Vocabulary v1.0.2 — Pre-Spike Materialization Contract (v1.0.2). Content seal SHA-256 5484226f9653792746e46edfbb2ae579b8d14ccce233525dbf197873ecb3a063.

Repository-authored inputs (not Drive exports)

SocioLens claim — LD04-SL-C06 v1.0 (v1.0) — written in the SocioLens repository by Project Lead (Claude, AI) on 2026-10-14; no Drive identity exists. Authority: AGENTS.md "Authority" (the Project Lead decides research content within the research-integrity rules) and the Owner's choice in the SocioLens project chat on 14 October 2026 to build the LD-04 debate; a new Claim, first version, written from the project's LD-04 source notes, checked against full texts or official abstracts. Block “ENTRY 06 — LD04-SL-C06” SHA-256 dba22541fda94585e64b47187ea35ad12472f89ddd920f0aa8aea72ec494603e; fixture fixtures/authored/claims/LD04-SL-C06--v1.0.

Builder decisions (not Drive data)

VERSION = v1.0 — basis: Claim version taken from the repository-authored claim fixture version (SocioLens claim — LD04-SL-C06 v1.0); the first version of a new Claim of the LD-04 debate
CURRENT_POINTER = true — basis: repository-authored claim "SocioLens claim — LD04-SL-C06 v1.0" (Project Lead (Claude, AI), 2026-10-14); basis: AGENTS.md "Authority" (the Project Lead decides research content within the research-integrity rules) and the Owner's choice in the SocioLens project chat on 14 October 2026 to build the LD-04 debate; a new Claim, first version, written from the project's LD-04 source notes, checked against full texts or official abstracts
FIELD_6_SOURCE_REFERENCE = SRC-SL-IDAN-ANAND-ARXIV-2605-18143--v1.0 — basis: field 6 of ENTRY 06 — LD04-SL-C06 cites "Idan & Anand, “Generative AI and the Productivity Divide: Human-AI Complementarities in Education,” arXiv:2605.18143, May 2026"; the only field-6 segment of this ClaimVersion's repository-authored ledger block.
SOURCE_PROVENANCE_PIN = SRC-SL-IDAN-ANAND-ARXIV-2605-18143--v1.0 — basis: explicit version-pinned public provenance of this ClaimVersion (Amendment v1.2 §6): the pinned field-6 Source record "Idan & Anand, “Generative AI and the Productivity Divide: Human-AI Complementarities in Education,” arXiv:2605.18143, May 2026".