I came here to comment because this was posted by… Bender, which I found hilarious.

It’s not even the key combo, iOS and autocorrect will do it for you.

Yep. Like I said elsewhere on the thread, some of them already have enough karma to downvote.

Yeah, and some of them already have enough karma to downvote you if you call them out, which is infuriating…

I still call voodoo on this. I use an iPhone, iPad, Mac to comment here—all of them autocorrect to em dashes at one point or another. Same goes for ellipsis.

Why would recently created accounts be 10x more likely to be created by owners of Apple products or English majors than the baseline?

I doubt it explains any reasonable fraction of this, but github moving from early adopter techies to general population "normies" would be a reason for the shift. I would expect it explains at least some increase in the use of em-dashes.

Do general population normies really use em-dash, or do they just reach for the dash they see clearly printed on their keyboard?

I think they're pressing the default dash (actually a hyphen) twice, and that autocompletes to a single em dash.

You can remove em dashes from the analysis and the trend is still there: newly created accounts are still 6X more likely to use the remaining LLM indicators (arrows and bullets, p = 0.00027).

Ellipses were never part of the analysis.

I’ve built my own flavor (https://github.com/rcarmo/piclaw) and use it to batch convert documentation, check some news (it runs Playwright and sends me summaries of specific sites that don’t have RSS feeds) and do moderate tweaks to my website (like re-tagging posts and adding reference links).

Thanks for sharing!

Nice trick. I’m going to see how I can apply it to tool calls in pi.dev as well

That means a lot, thank you! Would love to hear your feedback once you try it — and an upvote would be much appreciated if you find it useful

I run it almost exclusively with codex models. Zero issues.

Paraphrasing The Dude, that’s like, just your opinion, man.

I run mine inside https://github.com/rcarmo/agentbox (with https://github.com/rcarmo/webterm)