Announcement The use of AI

[Raigryn Vayd]

Alright everyone, let's have that discussion about AI LLMs (e.g. ChatGPT).

It's something that has become a lot more pervasive and normalised in our daily lives.

Here at Chronicles we do embrace new technology, but we want to have a discussion about what's appropriate for it's use.

This is a creative hobby. We come here to collaboratively write and make new friends. I always ask myself: would I enjoy simply writing backwards and forwards with an LLM?

The answer is invariably no.

No one expects perfect writing. The important part is that you're participating and enjoying this. Your writing partner deserves a written response, not something automatically generated.

Let's start with the non-exhaustive list of where the use of AI is appropriate (it's quite long):

• Generating art
• Helping with biographies and templates
• Coming up with roleplay ideas and names
• Grammar, spelling checks and translations between languages

What is not okay:

• Generating your roleplay post.

There, that's quite simple really.

Are we enforcing this?

No. At least, not yet. If we felt there was someone truly flaunting this we would start with an open conversation and go from there.

But by setting out this policy we set our expectations on people collaborating in our roleplay community.

Attempting to sell AI generated images as your own has already resulted in a Ban. Definitely do not, under any circumstances, do this.

We also will not tolerate witch hunts. If you think something is AI, report it for us to review. If you harass someone that you think is posting AI content, you are also violating our rules.

If you have any questions or concerns, please raise them here, but please keep all conversation respectful. This is also not a discussion about AI ethics. For clarity, don't bring that to this thread.

 

[demonz-r-us]

Well-said. I do have a question about what you said would be an appropriate use of AI. You did mention that writing a post in AI is unacceptable, but would having AI tweak a few things and set the tone of the overall roleplay post be acceptable? E.g
Original: "Bob walks into the tavern with a steady pace, hair clinging to his forehead in clumps of sticky sweat. His eyes dart between inhabitants before he swiftly settles down into a chair, ordering three shots of beer with a glance at the bartender before downing all three in quick succession."
AI edited: "Bob pushed through the tavern door with a slow, deliberate stride, the kind that suggested a long day gnawing at his heels. Strands of damp hair clung stubbornly to his forehead, shining with the sheen of stale sweat. His gaze swept across the room—past the rowdy dice players, the murmuring travelers, the lone musician tuning a battered lute—before he slipped into an empty chair as if trying not to disturb the air around him. Without a word, he lifted his chin toward the bartender, signaling for a round of drinks, his expression unreadable beneath the flickering lantern light. Quickly and smoothly, the round is downed before he has a chance to catch up with his own thoughts."

I ask this question because I do use AI in my posts to help me better the atmospheric tone. Of course, now that a "warning flag" has been raised, I'll probably stop doing even that to insure the authenticity of my writing is kept in check. Wouldn't want to get in a habit if it's against the rules in the near - or even far - future.

 

[Raigryn Vayd]

demonz-r-us

I'm going to speak from the gut here.

That second version is so absolutely, utterly and obviously AI written that anyone would spot it from a mile. If you hand wrote that to look like ChatGPT, that's impressive.

The first post is a genuinely better paragraph for a roleplay thread.

I don't feel I can add much more than that.

 

[Darkweaver]

Interesting, I didn't even consider it might be prevalent or show up much in roleplay posts. But I suppose it's easy to do.

Personally I refuse to let AI take over one of the few skills I feel I've developed over a long period. I use it only for art, since I can't draw to save my life. It might be delusional, but I hope to at least beat the machine in writing.

 

[Hrefna]

As someone who's written on forums since overtaken by AI this is such a refreshing take and something I'll remember for the rest of my stay here!

If I can add to the discussion above(for those who may report generated posts), there are folks like myself who paste Em dashes—usually to inject thoughts and "fun facts"—into our posts. This is sometimes mistaken as a sign of AI generation. It can be I think? but only when it's used excessively throughout a post in my experience. I think someone mentioned journalists using it a lot and AI stealing it from their articles.

Anyway, I'm glad to see AI enshitification of the hobby is taken seriously here lol

 

[Velaeri]

As someone who's written on forums since overtaken by AI this is such a refreshing take and something I'll remember for the rest of my stay here!

If I can add to the discussion above(for those who may report generated posts), there are folks like myself who paste Em dashes—usually to inject thoughts and "fun facts"—into our posts. This is sometimes mistaken as a sign of AI generation. It can be I think? but only when it's used excessively throughout a post in my experience. I think someone mentioned journalists using it a lot and AI stealing it from their articles.

Anyway, I'm glad to see AI enshitification of the hobby is taken seriously here lol

I work in marketing and content generation as a career. Em dashes are definitely one of the many key indicators of AI written work, but are not an end-all-be-all of it. I also like using dashes, though I hardly put in the effort to do a full em dash.

LLMs are trained off human writing. Humans used em dashes first, but the people who use them usually do so with purpose and intent, irregularly, not like sprinkles on ice cream cones that we see from LLMs.

In general, those who work with LLMs on the regular (like myself) can spot AI generated content from ten miles away. We know the giveaways. This is why we are asking for members to report content they feel may be breaking this rule rather than calling it out publicly or harassing the author directly.

It's not always cut and dry. It usually requires a discussion or two.