# I am almost certain the “It’s not ___, it’s ___” AI writing style came from Chinese — retained record

> Retained reddit record from reddit:ChatGPT.

Canonical: https://highsignal.app/data/records/4ff3632bbcd0fd3e

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:ChatGPT)  published 2026-10-20 23:09 ingested 2026-10-21 02:36

[original source ↗](https://www.reddit.com/r/ChatGPT/comments/1xb73pu/i_am_almost_certain_the_its_not_its_ai_writing/)

## retained text



```
A classic telltale sign of AI writing is the “it’s not…, it’s…” pair, but a lot of people are confused (me included) by where this came from. It’s not a very popular way to construct sentences in English, so it’s weird it became such a hallmark of AI writing. The other interesting thing is the pair produced by AI is rarely an antithesis or that different in meaning. For example, you’d expect a sentence starting with “it’s not a discovery, it’s…” to end with something in opposition, like “it’s a well known fact”, not “it’s not a discovery, it’s a journey of exploration“. A journey of exploration is in essence the same as a discovery, making the sentence sound like an unnecessary tautology. No one with English as their first language writes like that. 

However, in Chinese, there is a roughly equivalent pair 不是，而是 that is extremely common and widely used. I remember having to construct sentences using this pair as part of my primary school homework growing up. It’s absolutely ubiquitous in Chinese writing. The one crucial difference between the two is that this pair is used to clarify nuances or add to sentiments, not literally for opposites.

If you translated the above example into Chinese (不是一次发现，而是一段探索之旅), it reads very much like generic marketing copy but makes far more sense than the English equivalent. It’s absolutely believable that some person wrote that. The more correct sounding English version would actually be considered an incorrect use of 不是，而是.

Additionally, I feel like this sentence construction popped up about the same time as articles about how western models now tend to believe they are Chinese, due to using Chinese models to train. 

I feel like I have discovered something here. Thoughts?
```



document fetched 2026-10-21 00:20 · canonical https://reddit.com/r/ChatGPT/comments/1xb73pu/i_am_almost_certain_the_its_not_its_ai_writing
