Most active commenters
  • Moru(3)

←back to thread

Show HN: Hacker News em dash user leaderboard pre-ChatGPT

(www.gally.net)
358 points tkgally | 18 comments | 30 Aug 25 03:40 UTC | HN request time: 0.771s | source | bottom

The use of the em dash (—) now raises suspicions that a text might have been AI-generated. Inspired by a suggestion from dang [1], I created a leaderboard of HN users according to how many of their posts before November 30, 2022—that is, before the release of ChatGPT—contained em dashes. Dang himself comes in number 2—by a very slim margin.

Credit to Claude Code for showing me how to search the HN database through Google BigQuery and for writing the HTML for the leaderboard.

[1] https://news.ycombinator.com/item?id=45053933

1. PUSH_AX ◴[30 Aug 25 06:40 UTC] No.45072409[source]▶
It might be more fun to see users who’s emdash usage increased after the release.
replies(6): >>45072489 #>>45072739 #>>45072901 #>>45073019 #>>45073342 #>>45074025 #
2. Moru ◴[30 Aug 25 06:56 UTC] No.45072489[source]▶
Maybe the HN crowd is the wrong group for such statistics, a higher percentage here probably knows how to use their keyboard and OS.
replies(4): >>45072523 #>>45072595 #>>45072971 #>>45073083 #
3. 9rx ◴[30 Aug 25 07:18 UTC] No.45072595[source]▶
Plus being nerdier in general. I, for one, purposely use it more often because of all the hoopla.
replies(1): >>45072920 #
4. Moru ◴[30 Aug 25 07:48 UTC] No.45072739[source]▶
I missed the point of the leaderboards completely. It is to show exactly that when you get blamed for using AI to write. You can point out that you already used it in 2009 or whatever. For that it is very useful yes :-)
5. montebicyclelo ◴[30 Aug 25 08:19 UTC] No.45072901[source]▶
Although note — people are likely to be infuenced by the recent prevalence of em dash to use it more in their own writing nowadays
6. firesteelrain ◴[30 Aug 25 08:24 UTC] No.45072920{3}[source]▶
Burn him at the stake!
7. dns_snek ◴[30 Aug 25 08:31 UTC] No.45072971[source]▶
I think they meant after the release of ChatGPT. If someone never used them before and now uses them all the time it might indicate that they're using ChatGPT... or it might just mean that they learned how to use them after widespread discussions about it.
replies(2): >>45073595 #>>45074993 #
8. idiotsecant ◴[30 Aug 25 08:43 UTC] No.45073019[source]▶
Even more interesting is the likely increase in emdash usage by those not using an LLM, but merely imitating the writing they see subconsciously. There was a evidence that chatgpt is shifting the frequency of use of some uncommon words and phrases amongst non-users.
replies(1): >>45073327 #
9. perihelions ◴[30 Aug 25 08:54 UTC] No.45073083[source]▶
I remember participating in a small thread on how to type an em-dash, on different OS's. It was in March 2023, so before the em-dash meme had started—it was an innocent question then.

https://news.ycombinator.com/item?id=35118338#35118598

10. sebastiennight ◴[30 Aug 25 09:46 UTC] No.45073327[source]▶
Oh really? We should definitely delve into this.
replies(2): >>45073473 #>>45077944 #
11. dns_snek ◴[30 Aug 25 09:49 UTC] No.45073342[source]▶
HN is burying my comments (thanks!) but here it is: https://news.ycombinator.com/item?id=45073287
12. JdeBP ◴[30 Aug 25 10:22 UTC] No.45073473{3}[source]▶
You'll need to delve into history back quite a number of years. (-:

* https://news.ycombinator.com/item?id=18439869

13. withinboredom ◴[30 Aug 25 10:48 UTC] No.45073595{3}[source]▶
I use em-dashes now more than ever — mostly just to mess with people.
replies(1): >>45074403 #
14. akoboldfrying ◴[30 Aug 25 12:25 UTC] No.45074025[source]▶
Agreed.

More generally any measurable feature of writing that underwent a significant change in frequency around that time would be interesting to look at. Looking at frequencies across the entire post dataset would suggest likely candidates, which individual people could then be tested against. There would be lots of confounding factors and red herrings though -- like the word "ChatGPT" itself!

15. brookst ◴[30 Aug 25 13:19 UTC] No.45074403{4}[source]▶
Certainly, it’s great fun to trigger the AI skeptics.
replies(1): >>45081717 #
16. ◴[30 Aug 25 14:30 UTC] No.45074993{3}[source]▶
17. ◴[30 Aug 25 20:57 UTC] No.45077944{3}[source]▶
18. Moru ◴[31 Aug 25 09:13 UTC] No.45081717{5}[source]▶
It's not AI skeptics, it's users that does not know how to type — and is vulnerable to hype.