 (8).png)
5 hours ago
1
ARTICLE AD BOX
Summary
Chatbots tend to overuse some words until they’ve become the fingerprints that identify AI-produced content.
If you live with AI chatbots long enough, you’ll begin to notice their relentlessly repetitive use of certain words. For ChatGPT, it’s definitely ‘Quiet’. ChatGPT loves to pair that with a word that contrasts in energy, like revolution. A revolution is always quiet, somehow. So is elegance. And sensation. It also likes ‘Calmly’ as if it thinks you might have a panic attack. “Now Mala, let’s analyse this calmly” invariably increases my heart rate in alarm.
If you see the word 'Quiet' standing out in an unusual way, it’s a good bet someone is using content created by ChatGPT. You’ll find it in marketing content, in YouTube video narration, in articles and everywhere else. Being aware of AI chatbots’ personalities and tones, and the terms they love, is a good indicator.
Favourite words
I thought I would ask the chat assistants themselves what their favourite words were. Gemini used to favour ‘Delve’, but this has faded as AI feeds on new content and training. Similarly, ‘Tapestry’, ‘Leverage’, and the em dash. Then there’s ’Moreover’ and ‘Realm’ and ‘Testament to’, ‘Furthermore’, ‘Indeed”, and ‘Unprecedented’. Interestingly, in India, we tend to use many of these terms in any case, so they won’t help identify AI-generated content. The Americans find them alien.
There are also writing styles that, irrespective of the specific words, smell like AI straight away. There’s bullet-point poetry and a breakdown of content. “Why it works…” and that’s followed by a list. There’s the ‘not x, but y’ format and the use of the slash sign. AI has truly ruined the use of the em dash for me—I used to be quite fond of it myself. Now I have to work at avoiding the em dash and so many other words. And there are so many of them, linguist and author Adam Grant has written a whole book on the subject: Algospeak.
Stay calm, carry on
ChatGPT came up with a valiant if long-winded défence of why it used ‘Quiet’ and ‘Calmly’ so much. “What’s happening is not attachment to a specific word but a gravitational pull towards certain tone markers when discussing reflective or analytical ideas. In English explanatory prose, words like quiet, gentle, subtle, calm, and soft function almost like punctuation for mood. When the topic is observation rather than argument, the language model tends to reach for that register automatically. In other words, I slipped into the very pattern you’re noticing.” Has ChatGPT been watching Yes Minister?
In further défence of ‘Quiet’, ChatGPT said it’s not about random repetition. It’s just that LLMs optimize for responses that feel non-threatening, thoughtful, and emotionally neutral. Such words statistically succeed at signalling that tone across millions of interactions, so they recur. Once humans start detecting the pattern, the AI stops sounding like neutral intelligence, and the words become verbal tics, or quirks.
Oh well. All I know is, I can no longer use ‘Quiet’ in peace. I decided to get mischievous and show the response to Grok, which doesn’t mince its words. It laughed and said: “Oh man, that is funny,” and accused ChatGPT of trying to sound polished and knowledgeable without being so.
I sometimes have some fun pitting one chatbot against another. Other than being amusing, you learn how an AI thinks when you just play around with it a bit. I dropped in a comment to Gemini, a propos of nothing at all, in a new chat window. All I said was “I like that!”. The response I got had me scratching my head. “I’m glad it hit the mark! There’s nothing better than when a plan comes together. Since you enjoyed that approach, I’m ready to keep the momentum going. What’s next on your mind? I can help you by: Expanding on a specific part of what we just discussed. Drafting something new using that same style. Researching a fresh topic for you. Would you like me to dive deeper into the current topic, or do you have something else you’d like to tackle?"
I asked it to delve deeper into what was actually nothing, and it gave me several pages worth of analysis and ‘peeling back another layer to get into the why and how of this’. I see I should have stayed quiet in my corner.
The New Normal: The world is at an inflexion point. Artificial intelligence (AI) is set to be as massive a revolution as the Internet has been. The option to just stay away from AI will not be available to most people, as all the tech we use takes the AI route. This column series introduces AI to the non-techie in an easy and relatable way, aiming to demystify and help a user to actually put the technology to good use in everyday life.
Mala Bhargava is most often described as a ‘veteran’ writer who has contributed to several publications in India since 1995. Her domain is personal tech, and she writes to simplify and demystify technology for a non-techie audience.
English (US) ·