In argumentative essays co-written with a language model, only the instruction-tuned model (InstructGPT) made essays more alike; the base model (GPT-3) did not, and the reduction in diversity came from the text the model contributed, while the users’ own text remained unaffected.

SocioLens project,

Claim LD09-SL-C14 (v1.0)

Research opportunities this claim motivates

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

Does handing work to the model cause both skill loss and sameness?The reduction in essay diversity came from the text the model contributed; users' own text was unaffected.
Does the convergence caused by writing with AI persist or compound over months of use?Homogenization came from the model's contributed text, not the users' own text, within a session.

Provenance and status

Claim type: Estimate
Publication status: Eligible for publication, with a qualifier
Verification state: Checked against the primary source. 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: Effects of current models, or on essay quality
Limitations: Small corpus; one genre; models from 2022; read in the arXiv version 3 (July 2024), posted after the ICLR 2024 acceptance; the proceedings version was not read
Source (as cited): Padmakumar & He, “Does Writing with Language Models Reduce Content Diversity?,” ICLR 2024, arXiv:2309.05196
Source type: Peer-reviewed conference paper (recorded as “Peer-reviewed conference paper”)
Institutional position: Independent academic (New York University)
Maturity of the evidence: Emerging evidence
Recorded date: 15 October 2026
Currentness / correction: This is the current version. No correction or concern is recorded on its sources.

Sources this version cites

Padmakumar & He, “Does Writing with Language Models Reduce Content Diversity?,” ICLR 2024, arXiv:2309.05196 (Peer-reviewed conference paper)
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 (QUALIFIED) (as printed: “ELIGIBLE WITH QUALIFIER”)
Verification state (coded): CONTENT_VERIFIED_PRIMARY
Source type (coded): PEER_REVIEWED_CONFERENCE_PAPER
Maturity (coded): EMERGING

#	Canonical field	Canonical value (verbatim)	Normalized (Claim Value Vocabulary v1.0.2)
1	CLAIM ID	LD09-SL-C14	—
2	PUBLIC-FACING CLAIM	In argumentative essays co-written with a language model, only the instruction-tuned model (InstructGPT) made essays more alike; the base model (GPT-3) did not, and the reduction in diversity came from the text the model contributed, while the users’ own text remained unaffected.	—
3	CLAIM TYPE	ESTIMATE	base_types: ESTIMATE
4	RESEARCH STRAND	B — Creative work, writing and ideas	—
5	MATURITY	EMERGING	states: EMERGING
6	SOURCE	Padmakumar & He, “Does Writing with Language Models Reduce Content Diversity?,” ICLR 2024, arXiv:2309.05196	source_ids: SRC-SL-PADMAKUMAR-HE-ICLR-2024
7	SOURCE TYPE	Peer-reviewed conference paper	classes: PEER_REVIEWED_CONFERENCE_PAPER
8	INSTITUTIONAL POSITION	Independent academic (New York University)	—
9	SCOPE	Argumentative essays on ten topics	—
10	DESIGN	Controlled writing study with three setups (solo, base model, instruction-tuned model)	—
11	SAMPLE / DATASET	300 essays by paid native English-speaking writers	—
12	MODEL / TOOL / VERSION	GPT-3 (davinci) and InstructGPT (2022)	—
13	COMPARATOR	Essays written alone	—
14	MAIN RESULT	Key-point homogenization (Rouge-L): solo 0.1536, GPT-3 0.1578, InstructGPT 0.1660, with only InstructGPT significantly different; lexical and key-point diversity also lower with InstructGPT; user-written text unaffected	—
15	INFERENTIAL TARGET	Where in human–AI writing homogenization comes from	—
16	DOES NOT LICENSE	Effects of current models, or on essay quality	—
17	VERIFICATION STATE	CONTENT VERIFIED	state: CONTENT_VERIFIED_PRIMARY · verified_by: UNKNOWN
18	LIMITATIONS	Small corpus; one genre; models from 2022; read in the arXiv version 3 (July 2024), posted after the ICLR 2024 acceptance; the proceedings version was not read	—
19	CONTRADICTORY / SUPPORTING SOURCES	Not applicable	—
20	GEOGRAPHY / LANGUAGE	Not checked / English	—
21	DISTRIBUTIONAL LEDGER	Not recorded	—
22	R01 / FIELD-MAP LINKS	LD-09, D3	—
23	WHAT WOULD CHANGE THIS CLAIM?	Replication with current models and tuning methods	—
24	PUBLICATION STATUS	ELIGIBLE WITH QUALIFIER	state: ELIGIBLE · qualifiers: QUALIFIED
25	RECORDED DATE	2026-10-15	—

Field 22 references resolved against Field Map v0.2: LD-09 (LIVING_DEBATE); D3 (DOMAIN)

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

Pinned Sources (field 6, version-specific)

“Padmakumar & He, “Does Writing with Language Models Reduce Content Diversity?,” ICLR 2024, arXiv:2309.05196” → Padmakumar & He (2024), “Does Writing with Language…” (SRC-SL-PADMAKUMAR-HE-ICLR-2024--v1.0) — Padmakumar & He, “Does Writing with Language Models Reduce Content Diversity?,” ICLR 2024, arXiv:2309.05196; class PEER_REVIEWED_CONFERENCE_PAPER (explicit public provenance pin)

Cited by Living Debate versions

Individual gain, collective narrowing? Does AI make science, creative work and shared knowledge more alike? (v0.1) LD-09

Cite this page

SocioLens project. (2026, October 15). In argumentative essays co-written with a language model, only the instruction-tuned model (InstructGPT) made essays more alike; the base model (GPT-3) did not, and the reduction in diversity came from the text the model contributed, while the users’ own text remained unaffected [Claim LD09-SL-C14, version v1.0]. SocioLens. https://sociolens-web.pages.dev/claims/ld09-sl-c14/v1.0/

@misc{sociolens_ld09_sl_c14_v1_0,
  author = {{SocioLens project}},
  title = {In argumentative essays co-written with a language model, only the instruction-tuned model (InstructGPT) made essays more alike; the base model (GPT-3) did not, and the reduction in diversity came from the text the model contributed, while the users’ own text remained unaffected.},
  year = {2026},
  date = {2026-10-15},
  howpublished = {\url{https://sociolens-web.pages.dev/claims/ld09-sl-c14/v1.0/}},
  note = {Claim LD09-SL-C14, 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 30317c6bd9729368d9654d920bffac7cc18d23c9e1dbfac4a953b0a5a4eecc8b.

Repository-authored inputs (not Drive exports)

SocioLens claim — LD09-SL-C14 v1.0 (v1.0) — written in the SocioLens repository by Project Lead (Claude, AI) on 2026-10-15; 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 15 October 2026 to build the LD-09 debate; a new Claim, first version, written from the project's LD-09 source notes, checked against full texts or official abstracts. Block “ENTRY 14 — LD09-SL-C14” SHA-256 1c431a4f32f569fa7048f475f8aabb2f29e93590ff9e940702d9f00ba94ce2e4; fixture fixtures/authored/claims/LD09-SL-C14--v1.0.

Builder decisions (not Drive data)

VERSION = v1.0 — basis: Claim version taken from the repository-authored claim fixture version (SocioLens claim — LD09-SL-C14 v1.0); the first version of a new Claim of the LD-09 debate
CURRENT_POINTER = true — basis: repository-authored claim "SocioLens claim — LD09-SL-C14 v1.0" (Project Lead (Claude, AI), 2026-10-15); 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 15 October 2026 to build the LD-09 debate; a new Claim, first version, written from the project's LD-09 source notes, checked against full texts or official abstracts
FIELD_6_SOURCE_REFERENCE = SRC-SL-PADMAKUMAR-HE-ICLR-2024--v1.0 — basis: field 6 of ENTRY 14 — LD09-SL-C14 cites "Padmakumar & He, “Does Writing with Language Models Reduce Content Diversity?,” ICLR 2024, arXiv:2309.05196"; the only field-6 segment of this ClaimVersion's repository-authored ledger block.
SOURCE_PROVENANCE_PIN = SRC-SL-PADMAKUMAR-HE-ICLR-2024--v1.0 — basis: explicit version-pinned public provenance of this ClaimVersion (Amendment v1.2 §6): the pinned field-6 Source record "Padmakumar & He, “Does Writing with Language Models Reduce Content Diversity?,” ICLR 2024, arXiv:2309.05196".