Claims Library Entry
Your Coworkers Are Talking like ChatGPT and You Don't Even Know It
New research analyzing 740,000 hours of YouTube talks and podcasts shows that ChatGPT's favorite words like 'delve' and 'meticulous' are increasingly appearing in spontaneous human speech since the tool's release. The article explores how AI language patterns create a feedback loop that could homogenize workplace communication and reduce creative thinking. It offers practical steps teams can take to preserve their unique voice while using AI tools.
Published August 1, 2025 by Kamil Banc
Lead claim
New research shows ChatGPT's favorite words are measurably creeping into spontaneous human speech.
Atomic Claims
What this article supports
Copy individual claims as needed.
Claim 1 · Source summary
Massive Speech Study
A study analyzed 740,000 hours of human speech from YouTube talks and podcasts.
Claim 2 · Source summary
ChatGPT Words Rising
Researchers found sharp increases in words like delve, comprehend, boast, swift, and meticulous after ChatGPT's release.
Claim 3 · Source summary
Seven Billion Words
The dataset included 360,445 YouTube academic talks and 771,591 podcast episodes transcribed into over seven billion words.
Claim 4 · Source summary
GPT Score Methodology
Researchers measured ChatGPT word preferences by comparing human-written texts before and after ChatGPT edits, assigning GPT scores.
Claim 5 · Kamil's interpretation
Protecting Team Voice
Teams should document authentic communication patterns and require AI outputs to be translated into company voice.
Evidence
Context behind the claims
Quote
"Machines learned from human data, developed their own quirks, then taught those quirks back to us."
Key statistics
740,000 hours
Total human speech analyzed across YouTube talks and podcasts to detect AI language influence
360,445
YouTube academic talks included in the study dataset
771,591
Podcast episodes analyzed to test whether AI influence reaches spontaneous conversation
7 billion+
Words transcribed from the audio dataset to track shifts in word usage
Supporting context
The underlying study, published on arXiv, transcribed over seven billion words from 360,445 academic YouTube talks and 771,591 podcast episodes, then tracked usage of ChatGPT-preferred words before and after the tool's November 2022 release. Researchers built GPT scores by feeding human-written texts into ChatGPT and measuring which words appeared more often in its edits, and used synthetic controls to establish causation. For practitioners, the findings suggest auditing team communication for AI-influenced phrasing, documenting authentic speech patterns, and creating AI usage guidelines that require translating outputs back into the organization's natural voice. The research also offers an encouraging note: not every ChatGPT-preferred word gets adopted by humans, meaning teams retain agency over how AI shapes their language.
How to Cite
Use the claim-level citation when you need a precise statement. Use the article or claims-collection citation when you want the wider argument and source context.
Individual Claim
Best when you need to cite one atomic claim directly inside a memo, deck, research note, or AI output.
"[claim text]" (Banc, Kamil, 2025, https://kbanc.com/claims-library/your-coworkers-are-talking-like-chatgpt-and-you-dont-even-know-it)Original Article
Use this when you want to cite the full newsletter article at AI Adopters Club rather than the structured claims page.
Banc, Kamil (2025, August 1, 2025). Your Coworkers Are Talking like ChatGPT and You Don't Even Know It. AI Adopters Club. https://aiadopters.club/p/your-coworkers-are-talking-like-chatgptClaims Collection
Use this when you want to reference the full structured claims collection on this page.
Banc, Kamil (2025). Your Coworkers Are Talking like ChatGPT and You Don't Even Know It [Structured Claims]. Retrieved from https://kbanc.com/claims-library/your-coworkers-are-talking-like-chatgpt-and-you-dont-even-know-itAttribution Requirements
- Include the author name: Kamil Banc.
- Include the source: AI Adopters Club or the structured claims page.
- Link to the original article or the claims page you used.
- Indicate any edits or transformations if you changed the wording.
Related Reading
More from the library
Take-Two Interactive's CEO publicly claims AI has "no creativity" while the company files patents for advanced AI systems. This dual narrative protects a $12.7 billion AI strategy that includes automated world-building, AI-driven QA, and player behavior prediction engines acquired through Zynga.
5 claims
A handful of schools split work between AI-automated delivery and human judgment, compressing core curriculum into two focused hours. The remaining time opened for projects and face-to-face coaching, with students hitting mastery targets faster while teachers tripled mentoring time.
5 claims
Most AI rollouts fail despite extensive training because the real issue isn't capability—it's habit formation. This article reveals why 42% of AI initiatives were abandoned in 2025 and shows how to redesign workflows so AI becomes the path of least resistance, creating automatic adoption without force.
5 claims