When 43 researchers each spent over 100 hours carrying out randomly assigned ideas from the earlier study, review scores of the AI-generated ideas fell significantly more than those of expert ideas on novelty, excitement, effectiveness and overall, and on many measures human ideas then ranked higher.

SocioLens project,

Claim LD09-SL-C20 (v1.0)

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: A general verdict on AI in research
Limitations: Preprint; small sample; one field; same authors as the earlier study; the AI ideas carried out were generated by the earlier study’s agent, which runs on an Anthropic model, and this debate was drafted by an Anthropic model
Source (as cited): Si, Hashimoto & Yang, “The Ideation–Execution Gap: Execution Outcomes of LLM-Generated versus Human Research Ideas,” arXiv:2506.20803, June 2025
Source type: Preprint (recorded as “Preprint”)
Institutional position: Independent academic (Stanford 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

Si, Hashimoto & Yang, “The Ideation–Execution Gap: Execution Outcomes of LLM-Generated versus Human Research Ideas,” arXiv:2506.20803, June 2025 (Preprint)
Read the original

Version

This page shows version v1.0. Its record never changes, even when newer evidence arrives; the site may add a note to the page or withhold a figure not cleared for public use.

This pinned version is also the current version.

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	LD09-SL-C20	—
2	PUBLIC-FACING CLAIM	When 43 researchers each spent over 100 hours carrying out randomly assigned ideas from the earlier study, review scores of the AI-generated ideas fell significantly more than those of expert ideas on novelty, excitement, effectiveness and overall, and on many measures human ideas then ranked higher.	—
3	CLAIM TYPE	ESTIMATE	base_types: ESTIMATE
4	RESEARCH STRAND	B — Creative work, writing and ideas	—
5	MATURITY	EMERGING	states: EMERGING
6	SOURCE	Si, Hashimoto & Yang, “The Ideation–Execution Gap: Execution Outcomes of LLM-Generated versus Human Research Ideas,” arXiv:2506.20803, June 2025	source_ids: SRC-SL-SI-HASHIMOTO-YANG-ARXIV-2506-20803
7	SOURCE TYPE	Preprint	classes: PREPRINT
8	INSTITUTIONAL POSITION	Independent academic (Stanford University)	—
9	SCOPE	NLP research ideas carried out as short papers	—
10	DESIGN	Randomized execution study followed by blind review of the resulting projects	—
11	SAMPLE / DATASET	43 expert researchers	—
12	MODEL / TOOL / VERSION	Ideas generated by the agent in LD09-SL-C19	—
13	COMPARATOR	Expert-written ideas executed in the same way	—
14	MAIN RESULT	Scores of AI ideas “decrease significantly more than expert-written ideas on all evaluation metrics” (p < 0.05); “a flip in rankings where human ideas score higher” on many metrics	—
15	INFERENTIAL TARGET	Whether judged novelty at the idea stage survives execution	—
16	DOES NOT LICENSE	A general verdict on AI in research	—
17	VERIFICATION STATE	CONTENT VERIFIED PREPRINT	state: CONTENT_VERIFIED_PRIMARY · qualifiers: PREPRINT · verified_by: UNKNOWN
18	LIMITATIONS	Preprint; small sample; one field; same authors as the earlier study; the AI ideas carried out were generated by the earlier study’s agent, which runs on an Anthropic model, and this debate was drafted by an Anthropic model	—
19	CONTRADICTORY / SUPPORTING SOURCES	Judged novelty before execution: LD09-SL-C19	—
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?	Larger execution studies in other fields	—
24	PUBLICATION STATUS	ELIGIBLE WITH PREPRINT QUALIFIER	state: ELIGIBLE · qualifiers: PREPRINT
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)

“Si, Hashimoto & Yang, “The Ideation–Execution Gap: Execution Outcomes of LLM-Generated versus Human Research Ideas,” arXiv:2506.20803, June 2025” → Si et al. (2025), “The Ideation–Execution Gap: Execution…” (SRC-SL-SI-HASHIMOTO-YANG-ARXIV-2506-20803--v1.0) — Si, Hashimoto & Yang, “The Ideation–Execution Gap: Execution Outcomes of LLM-Generated versus Human Research Ideas,” arXiv:2506.20803, June 2025; class PREPRINT (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). When 43 researchers each spent over 100 hours carrying out randomly assigned ideas from the earlier study, review scores of the AI-generated ideas fell significantly more than those of expert ideas on novelty, excitement, effectiveness and overall, and on many measures human ideas then ranked higher [Claim LD09-SL-C20, version v1.0]. SocioLens. https://sociolens-web.pages.dev/claims/ld09-sl-c20/v1.0/

@misc{sociolens_ld09_sl_c20_v1_0,
  author = {{SocioLens project}},
  title = {When 43 researchers each spent over 100 hours carrying out randomly assigned ideas from the earlier study, review scores of the AI-generated ideas fell significantly more than those of expert ideas on novelty, excitement, effectiveness and overall, and on many measures human ideas then ranked higher.},
  year = {2026},
  date = {2026-10-15},
  howpublished = {\url{https://sociolens-web.pages.dev/claims/ld09-sl-c20/v1.0/}},
  note = {Claim LD09-SL-C20, 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 3c86343ff26e5fd0264ea61d9cdc70ab479a1eafb3f21e483556ea1350c6151d.

Repository-authored inputs (not Drive exports)

SocioLens claim — LD09-SL-C20 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 20 — LD09-SL-C20” SHA-256 ee8105ec6e0de4bc987916c32eee00d9f38f0ad821c2b903ea0cd46ddbd4f01c; fixture fixtures/authored/claims/LD09-SL-C20--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-C20 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-C20 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-SI-HASHIMOTO-YANG-ARXIV-2506-20803--v1.0 — basis: field 6 of ENTRY 20 — LD09-SL-C20 cites "Si, Hashimoto & Yang, “The Ideation–Execution Gap: Execution Outcomes of LLM-Generated versus Human Research Ideas,” arXiv:2506.20803, June 2025"; the only field-6 segment of this ClaimVersion's repository-authored ledger block.
SOURCE_PROVENANCE_PIN = SRC-SL-SI-HASHIMOTO-YANG-ARXIV-2506-20803--v1.0 — basis: explicit version-pinned public provenance of this ClaimVersion (Amendment v1.2 §6): the pinned field-6 Source record "Si, Hashimoto & Yang, “The Ideation–Execution Gap: Execution Outcomes of LLM-Generated versus Human Research Ideas,” arXiv:2506.20803, June 2025".