In an online experiment with 1,174 Argentine adults, AI access shrank the performance gap between education groups on a business problem-solving task from 0.548 to 0.139 standard deviations; in a follow-up module without AI in the same session, a gap of 0.200 standard deviations remained.

SocioLens project,

Claim LD04-SL-C03 (v1.0)

Research opportunities this claim motivates

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

Do professionals trained with AI lose the skill without it a year later?A module done without AI immediately after an AI-assisted task showed no loss, and the authors call it a diagnostic rather than a test of skill formation.

Provenance and status

Claim type: Estimate
Publication status: Eligible for publication, with qualifiers: preprint
Verification state: Checked against the primary source; the source is a preprint. 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: Conclusions about skill formation; the authors call the follow-up a diagnostic, not a general test of skill formation
Limitations: Short online task; the no-AI module came straight after the task, so it does not measure learning over time; also issued as NBER Working Paper 34851 (February 2026), whose figures may differ
Source (as cited): Cruces, Fernández Meijide, Galiani, Gálvez & Lombardi, “Does generative AI narrow education-based productivity gaps? Evidence from a randomized experiment,” arXiv:2608.04198, submitted 4 August 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

Cruces, Fernández Meijide, Galiani, Gálvez & Lombardi, “Does generative AI narrow education-based productivity gaps? Evidence from a randomized experiment,” arXiv:2608.04198, submitted 4 August 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 (PREPRINT)
Source type (coded): PREPRINT
Maturity (coded): EMERGING

#	Canonical field	Canonical value (verbatim)	Normalized (Claim Value Vocabulary v1.0.2)
1	CLAIM ID	LD04-SL-C03	—
2	PUBLIC-FACING CLAIM	In an online experiment with 1,174 Argentine adults, AI access shrank the performance gap between education groups on a business problem-solving task from 0.548 to 0.139 standard deviations; in a follow-up module without AI in the same session, a gap of 0.200 standard deviations remained.	—
3	CLAIM TYPE	ESTIMATE	base_types: ESTIMATE
4	RESEARCH STRAND	B — Skill after the tool is removed	—
5	MATURITY	EMERGING	states: EMERGING
6	SOURCE	Cruces, Fernández Meijide, Galiani, Gálvez & Lombardi, “Does generative AI narrow education-based productivity gaps? Evidence from a randomized experiment,” arXiv:2608.04198, submitted 4 August 2026	source_ids: SRC-SL-CRUCES-ET-AL-ARXIV-2608-04198
7	SOURCE TYPE	Preprint	classes: PREPRINT
8	INSTITUTIONAL POSITION	Independent academic	—
9	SCOPE	Adults aged 25–45 from three Argentine survey panels; September–November 2025	—
10	DESIGN	Randomized online experiment with an incentivized task of about 21 minutes, followed immediately by a module without AI	—
11	SAMPLE / DATASET	1,174 participants	—
12	MODEL / TOOL / VERSION	Generative-AI assistant provided in the task interface	—
13	COMPARATOR	Participants without AI access	—
14	MAIN RESULT	Education gap 0.548 SD without AI vs 0.139 SD with AI; AI raised everyone’s performance; in the follow-up without AI, treated participants did no worse than controls, lower-education treated participants scored about 0.171 SD higher, and a 0.200 SD education gap remained	—
15	INFERENTIAL TARGET	Education-based performance gaps with and immediately after AI assistance	—
16	DOES NOT LICENSE	Conclusions about skill formation; the authors call the follow-up a diagnostic, not a general test of skill formation	—
17	VERIFICATION STATE	CONTENT VERIFIED PREPRINT	state: CONTENT_VERIFIED_PRIMARY · qualifiers: PREPRINT · verified_by: UNKNOWN
18	LIMITATIONS	Short online task; the no-AI module came straight after the task, so it does not measure learning over time; also issued as NBER Working Paper 34851 (February 2026), whose figures may differ	—
19	CONTRADICTORY / SUPPORTING SOURCES	Fits compression while assisted; the immediate follow-up finds no loss, unlike same-session exams in school mathematics	—
20	GEOGRAPHY / LANGUAGE	Argentina / Not checked	—
21	DISTRIBUTIONAL LEDGER	Lower-education participants gain most while assisted; the gap only partly closes underneath	—
22	R01 / FIELD-MAP LINKS	LD-04, D2, L-A	—
23	WHAT WOULD CHANGE THIS CLAIM?	A version change in the figures, or a delayed follow-up that shows a different pattern	—
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); L-A (LENS)

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

Pinned Sources (field 6, version-specific)

“Cruces, Fernández Meijide, Galiani, Gálvez & Lombardi, “Does generative AI narrow education-based productivity gaps? Evidence from a randomized experiment,” arXiv:2608.04198, submitted 4 August 2026” → Cruces et al. (2026), “Does generative AI narrow…” (SRC-SL-CRUCES-ET-AL-ARXIV-2608-04198--v1.0) — Cruces, Fernández Meijide, Galiani, Gálvez & Lombardi, “Does generative AI narrow education-based productivity gaps? Evidence from a randomized experiment,” arXiv:2608.04198, submitted 4 August 2026; class PREPRINT (explicit public provenance pin)

Cited by Living Debate versions

Cite this page

SocioLens project. (2026, October 14). In an online experiment with 1,174 Argentine adults, AI access shrank the performance gap between education groups on a business problem-solving task from 0.548 to 0.139 standard deviations; in a follow-up module without AI in the same session, a gap of 0.200 standard deviations remained [Claim LD04-SL-C03, version v1.0]. SocioLens. https://sociolens-web.pages.dev/claims/ld04-sl-c03/v1.0/

@misc{sociolens_ld04_sl_c03_v1_0,
  author = {{SocioLens project}},
  title = {In an online experiment with 1,174 Argentine adults, AI access shrank the performance gap between education groups on a business problem-solving task from 0.548 to 0.139 standard deviations; in a follow-up module without AI in the same session, a gap of 0.200 standard deviations remained.},
  year = {2026},
  date = {2026-10-14},
  howpublished = {\url{https://sociolens-web.pages.dev/claims/ld04-sl-c03/v1.0/}},
  note = {Claim LD04-SL-C03, 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 b0008feff7d1d28ccf1214773226da3bae48e2f399da4e00f3a54278c5e26f94.

Repository-authored inputs (not Drive exports)

SocioLens claim — LD04-SL-C03 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 03 — LD04-SL-C03” SHA-256 4549aabdccdec5b8845669b357af755bc132c1909140dd32913b9e6eb7120a90; fixture fixtures/authored/claims/LD04-SL-C03--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-C03 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-C03 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-CRUCES-ET-AL-ARXIV-2608-04198--v1.0 — basis: field 6 of ENTRY 03 — LD04-SL-C03 cites "Cruces, Fernández Meijide, Galiani, Gálvez & Lombardi, “Does generative AI narrow education-based productivity gaps? Evidence from a randomized experiment,” arXiv:2608.04198, submitted 4 August 2026"; the only field-6 segment of this ClaimVersion's repository-authored ledger block.
SOURCE_PROVENANCE_PIN = SRC-SL-CRUCES-ET-AL-ARXIV-2608-04198--v1.0 — basis: explicit version-pinned public provenance of this ClaimVersion (Amendment v1.2 §6): the pinned field-6 Source record "Cruces, Fernández Meijide, Galiani, Gálvez & Lombardi, “Does generative AI narrow education-based productivity gaps? Evidence from a randomized experiment,” arXiv:2608.04198, submitted 4 August 2026".