--- title: "People started saying \"delve\". Then they stopped." url: https://deadwritingtheory.com/cases/delve-out-loud/ summary: "ChatGPT's favourite words rose in unscripted podcast speech, \"delve\" to about 44% above expected. Then use fell below expected once people noticed. A preprint." published: 2026-10-09 updated: 2026-10-09 author: "Jack Stovell" publisher: "Adapt Progress Evolve Limited" language: en-GB --- # People started saying "delve". Then they stopped. ChatGPT's favourite words rose in unscripted podcast speech, "delve" to about 44% above expected. Then use fell below expected once people noticed. A preprint. **Case No. DWT-013: Delve, out loud.** Split verdict. Filed 9 October 2026. - **The question:** Did words that ChatGPT favours become more common in spontaneous human speech after its release, and did that last? - **The sample:** 737,083 hours of conversation from 824,634 podcast episodes, screened for unscripted speech, compared with a synthetic control built from words ChatGPT does not favour; plus a preregistered experiment with 496 people - **The study:** Hiromu Yakura, Ezequiel Lopez-Lopez, Levin Brinkmann, Ignacio de la Serna, Lara Kirfel, Prateek Gupta, Ivan Soraperra, Thomas F. Eisenmann, Dirk U. Wulff, Iyad Rahwan (Max Planck Institute for Human Development), "Empirical evidence of Large Language Model's influence on human spoken communication", arXiv preprint (version 4, July 2026), 2026. https://arxiv.org/abs/2409.01754 - **Whose study:** Third-party study. Not our study. Entered into evidence from Hiromu Yakura, Ezequiel Lopez-Lopez, Levin Brinkmann, Ignacio de la Serna, Lara Kirfel, Prateek Gupta, Ivan Soraperra, Thomas F. Eisenmann, Dirk U. Wulff, Iyad Rahwan, Max Planck Institute for Human Development. - **Peer review:** Preprint, not yet peer reviewed ## Findings of fact 1. Words ChatGPT prefers, such as delve, showcase, boast, intricacies and meticulous, increased abruptly in spontaneous speech. 2. In Science and Technology podcasts, "delve" rose to about 44% above its expected level 13 to 18 months after ChatGPT's release. 3. "The elevation did not persist": after peaking around mid-2024, "delve" fell below its expected level, 15% below in Science and Technology, 30% below in Business and 35% below across all categories. 4. In the experiment (N = 496), a short chat with a bot that used particular words led people to use those words themselves after a distraction task. ## Caveats for the jury - A preprint, not yet peer reviewed. - English podcasts only: a self-selected, public-facing group of hosts and guests. - It is about speech, not writing, and a few favoured words rising and falling is not a measure of sameness. ## The court reporter's account This one's odd, because it's about speech, not writing, and that matters. You cannot copy and paste into a spontaneous conversation. If the machine's words turn up there, somebody absorbed them. The paper is "Empirical evidence of Large Language Model's influence on human spoken communication", an arXiv preprint, revised July 2026, not peer reviewed. The authors are Yakura, Lopez-Lopez, Brinkmann, de la Serna, Kirfel, Gupta, Soraperra, Eisenmann, Wulff and Rahwan, of the Max Planck Institute for Human Development. The method has three parts. First, 737,083 hours of conversation from 824,634 podcast episodes, screened for unscripted speech. Second, a synthetic-control comparison against words ChatGPT does not favour. Third, a preregistered experiment with 496 people. For the prosecution: words ChatGPT prefers (delve, showcase, boast, intricacies, meticulous) increased abruptly in spontaneous speech. In Science and Technology podcasts, "delve" rose to about 44% above its expected level 13 to 18 months after ChatGPT's release. And in the experiment, a short chat with a bot that used particular words led people to use those words themselves after a distraction task. So the bot's vocabulary got into people. That's the claim, and the evidence is decent. For the defence, and I'll quote the authors here: "the elevation did not persist". Use of "delve" peaked around mid-2024, then fell below its expected level. It sat 15% below in Science and Technology, 30% below in Business, and 35% below across all categories. The turn coincides with the free release of GPT-4o and growing public awareness of "delve" as a sign of AI language. The authors also note "delve" may become stigmatised. Which is a strange thing to say about a word, but there it is. The point for the court is simple. The machine's words do get into people, and once people notice them, they back away. Both halves are in the record. Now the caveats, because I'm bound to read them out. It's a preprint. It covers English podcasts only, a self-selected public-facing group, so don't stretch it to everyone's kitchen table. And lexical shifts are not a measure of sameness: a handful of favoured words rising and falling tells you about those words, not about whether speech as a whole became more alike. This is neither a clean prevalence study nor a clean convergence study. It's about contagion, and then about retreat. The prosecution gets its contagion. The defence gets its retreat. Verdict: split. ## The exhibits entered in this case ### For the prosecution - **Exhibit P-16**: "Delve" got into unscripted speech. **44%** above expected use of "delve" in Science and Technology podcasts. About 44%, 13 to 18 months after ChatGPT's release. 737,083 hours from 824,634 podcast episodes, screened for unscripted speech. ### For the defence - **Exhibit D-17**: Then people backed away from "delve". **35%** below expected use of "delve" across all podcast categories, after mid-2024. 15% below in Science and Technology, 30% below in Business. The turn coincides with the free release of GPT-4o and public awareness of "delve" as a sign of AI. ## The finding ChatGPT's favourite words rose in unscripted podcast speech, "delve" to about 44% above expected. Then use fell below expected once people noticed. A preprint. ## Related - [The machine's words spread through medicine, measured without a detector.](https://deadwritingtheory.com/cases/the-delve-study-done-properly/) - [Not an invention. An inheritance.](https://deadwritingtheory.com/cases/where-did-ai-isms-come-from/) ## Sources - [Yakura et al., "Empirical evidence of Large Language Model's influence on human spoken communication", arXiv preprint (version 4, July 2026), 2026 (the original study)](https://arxiv.org/abs/2409.01754) ## Questions people ask ### Do people now talk like ChatGPT? For a while, a little. Yakura et al. (Max Planck Institute for Human Development, arXiv preprint) found "delve" rose to about 44% above its expected level in unscripted Science and Technology podcast speech, 13 to 18 months after ChatGPT's release. ### Did it last? No. In the authors' words, "the elevation did not persist". After mid-2024, "delve" fell below its expected level: 35% below across all podcast categories. The turn coincides with the free release of GPT-4o and public awareness of "delve" as a sign of AI. ### Why does speech matter to a theory about writing? Because nobody pastes text into a live conversation. If the machine's words turn up there, people absorbed them. ### Is this peer reviewed? Not yet. It is an arXiv preprint, revised in July 2026.