Bryan Mitchell

@PromptedInk
165 Followers
533 Following
3.4K Posts
Writer, Anglophile, Nostalgia Geek. Power User sans excess apps. Meshed between the literary ether and some leftover code. BA English as of 2014.
Bloghttps://prompted.ink
Blog (Archived)https://promptedink.blogspot.com
Bluesky Unbridgedhttps://bsky.app/profile/prompted.ink
Goodreadshttps://www.goodreads.com/user/show/30701953-bryan-mitchell

If you're trying to understand the impact of mastodon.social's new Terms of Service (which will also be the template for new instances once 4.4 is released) ... well, you're not alone. Here's a few discussions that I've found useful.

@neil has a blog post with Some thoughts on mastodon.social's updated terms

@sarahjamielewis has a thread at https://mastodon.social/@sarahjamielewis/114699476927561899

@mcc looks at the lack of a termination clause at https://mastodon.social/@mcc/114699201989866226

And here's the actual text of the new Terms of Service - https://techcrunch.com/2025/06/17/mastodon-updates-its-terms-to-prohibit-ai-model-training/

#mastodon #fediverse

Some thoughts on mastodon.social's updated terms

mastodon.social - one of the larger fediverse servers - has updated terms of service.

Accusations of AI use are flying around — often based on nothing more than vibes. But there’s (almost) no shortcut to knowing for sure in any one case.

And em dashes? They’re not the enemy. They’re glorious.

Check out the whole show for all the details:

Watch: https://youtu.be/zM0QAxF5P1k

Read: https://grammar-girl.simplecast.com/episodes/ai-em-dash

Listen: https://pod.link/173429229

Are em dashes really a sign of AI writing? 'Caretaker' vs. 'caregiver.' Episode 1091

YouTube

Here’s what else I found:

— There is some research showing that AI tends to overuse certain words, like "delve" and "meticulous." But this is based on looking at hundreds of thousands of papers and can't really tell you anything about one piece of writing. Don't stop using these words!

— Software-based AI detectors are also quite bad at what they're supposed to do.

— Marginalized groups are far more likely to be falsely accused of having used AI in their writing, and this causes real harm.

OMG, em dashes are not a sign of AI writing!

I kept hoping this myth would burn itself out, but no — the idea that em dashes are a sign of AI writing just keeps spreading. I finally couldn't take it anymore, so I crashed a segment into the latest Grammar Girl podcast.

The whole thing seems to have taken off after a popular GenZ culture podcast advised people to stop using the “the ChatGPT hyphen.” (And I have to confess it kills me that they didn't even know the name for it.) 🧵

I thought #DrWho fans would like this
Also: big US newspapers (with, I'm sure, some exceptions), we see you. We have done for some time, but today especially. One day you will either be ashamed or will not even exist in any meaningful way.
I hope US folks on the No Kings rallies know how much people outside the US are heartened to see them. We always knew the US is full of very fine people who totally reject and revile what's happening, but my god it's good to see it.
Call me a “weirdo” but I’m not changing how I express myself to appease AI hall monitors who say “only LLMs use em-dashes”.

To me, writing is an artform—em-dashes are essential to my art.

Don’t fuck with my art.

Someone else brought this up, and I think it’s true:

The online fight over em-dashes is really about neurotypical people trying to enforce their preferred mode of communication unto neurodivergent people.

That is, if you don’t talk exactly like them, they’ll literally say you’re not human.

If people are going to censor em-dashes—something that’s objectively unharmful—then online spaces really do have a problem with freedom of expression.

An em-dash may seem inconsequential.

But its perceived inconsequence is precisely why it should be defended.

If the humble em-dash can’t be defended, what will happen when more consequential freedom of expression needs defending?