(169) The Silent Author? Measuring ChatGPT’s Early Impact on Sexual Medicine Scientific Writing
Bibliographic record
Abstract
Abstract Introduction The emergence of large language models (LLMs) such as ChatGPT has begun to influence academic writing across biomedical disciplines. Prior studies have identified measurable linguistic shifts following ChatGPT’s release, including increased usage of stylistic enhancer terms often associated with LLM-generated text. However, the extent to which generative AI tools have impacted scientific writing within sexual medicine and fertility literature remains unstudied. Objective To evaluate the early linguistic influence of ChatGPT on sexual medicine and fertility journal abstracts by quantifying deviations in ChatGPT-associated vocabulary through counterfactual trend analysis. Methods Abstracts published between 2010 and 2024 were retrieved from 15 major sexual medicine and fertility journals via PubMed (n = 34,471). After exclusion of empty entries, 29,670 abstracts were included for analysis. A predefined list of 118 ChatGPT-associated marker terms was applied. Annual frequencies from 2010–2020 were used to model expected term usage had ChatGPT not been introduced (counterfactual projection). Observed frequencies for 2023–2024 were compared to projections using excess frequency gap (absolute difference) and excess ratio (relative increase). Results Counterfactual modeling using 2010–2020 data established baseline expectations for ChatGPT-associated vocabulary. In 2024, among 118 tracked ChatGPT-associated terms, 39 terms exhibited ≥2-fold excess ratios, and 7 terms exceeded 4-fold increases relative to projection. The leading terms by excess ratio were unparalleled (9.98-fold), intricate (6.77), intersection (4.69), elevate (4.44), underscore (4.15), delve (4.08), and valuable (4.07). The top terms by absolute excess frequency included comprehensive (+5.24%), additionally (+4.18%), valuable (+3.15%), crucial (+3.11%), and explore (+2.29%). While absolute frequencies for some of these stylistic terms remain modest, several (e.g. comprehensive, valuable, crucial) have reached usage rates exceeding 4–10% of abstracts in 2024, reflecting their growing integration into the narrative style of recent publications. In parallel, the prevalence of abstracts containing ≥1 ChatGPT-associated term rose from 5–8% during 2010–2019 to 9.5% in 2023 and 9.7% in 2024. By contrast, common control phrases demonstrated only minor changes over time, supporting the specificity of the observed linguistic shifts. Conclusions This counterfactual analysis provides early evidence that generative AI tools are introducing recognizable stylistic shifts into sexual medicine and fertility academic writing. Although absolute prevalence remains modest relative to broader biomedical literature, the direction and timing of change parallel global trends seen in AI-assisted scientific writing. These results underscore the need for ongoing monitoring, ethical guidelines, and transparent disclosure regarding the use of LLMs in academic publishing. Disclosure No
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.051 | 0.327 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.005 | 0.006 |
| Science and technology studies | 0.001 | 0.003 |
| Scholarly communication | 0.004 | 0.004 |
| Open science | 0.001 | 0.004 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.006 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".