Ohki's Trash

Check this out.

I need every single person to understand how horrible tumblr’s tagging system is

I go into the tag for epilepsy and its all flashing lights. We can’t use our own tag because people without epilepsy fill it up with improper warnings.

Use ‘flashing’ in place of ‘epilepsy’ in your tags. You aren’t warning people of epileptics, you’re warning us of flashing lights. Please please tag properly. Epileptics say this endlessly and constantly and it’s ignored. You are risking lives by doing this.

Here’s proof of what I mean:

Some characters that I did for our future Icewind Dale campaign.

Please vote and reblog but please do not reblog with a comment in the text of your reblog

When you reblog please tag with the same FIRST tag as the person you're reblogging from

Feel free to reply that's not what I'm looking at and I feel like we need somewhere to yell.

Again, you're going to vote in the poll (answer with the count closest to the bottom of the post you're seeing), you're going to reblog without any text or comments to avoid making a new reblog thread but still check the spread of the post, you're going to tag the post with the same FIRST tag as the person you're reblogging from (feel free to add other tags if you want, but the first one should be exactly the same as the person you're reblogging from).

Vote for the reblog count closest to the bottom of the post you see

Reblog with no text in the body of the reblog

Tag with the same first tag as prev

Thank you for your participation in this experiment!

New MaoMao and Jinshi sticker coming soon!

Now let me be clear - Gen AI is awful and most people agree that it has no place in creative spaces, but that's besides the point. I'm not here to debate AI in creative spaces, I'm here to talk about the moderators who feel the need to investigate and punish writers who they believe use AI.

I'm going to start by reminding people of two important things:

There is no way for anyone to know for certain if someone is using AI to write - unless they tell you they are. The common "tells" that are cited online are found in the work of people who have excellent control of functional grammar, good vocabularies, and an excellent control of sentence structure. AI is trained on the work of talented writers.

There is no way for an author to defend themselves against a claim that they are using AI, especially in an online environment. Nor should an author have to. The rule should not be "prove you are innocent or we will find you guilty!"

I feel very passionately about this and have for quite some time. I've seen fandom witch hunts before, and I know how easy it is for some people to climb to moral high ground and punish those they believe are doing the wrong thing.

My friend recently chose to leave a discord server after the mod team contacted them to announce that there had been reports made by people who wished to be anonymous who suspected their work was AI generated. The mods explained that they had investigated these claims by looking over their writing, and although they admitted they could not be 100% certain the work contained AI, they were issuing a strike/warning to them in accordance with their discord rules because they believed they might have.

The investigation involved them reading some of their work. 

That was it. 

They read their work and decided that, in their own words, they thought it might be AI so they were issuing a strike. They then proceeded to delete the writing they had shared on the server without any opportunity for this person to retrieve it.

There was no communication prior to them issuing a strike. My friend was not told about these accusations ahead of time, the moderators did not talk to them about the accusations or raise concerns, and did not stop to ask themselves why the people making these claims wished to do so anonymously.  

Perhaps if they had, they would have learned that members of their discord had been harassing this person via Tumblr anons about this issue. Instead they chose to give the bullies anonymity, to deny a writer a chance to have open communication, and issued a punishment for a rule infraction that they themselves admitted they could not be certain had been broken.

When my friend gave the names of beta readers and editors who had worked with her in real time, these moderators chose not to engage with them or seek clarification, and instead doubled down and tried to justify their decision.

My friend asked for the evidence they had that their work contained AI content, and it was not provided to them. Instead they supplied a generic statement about how they had made their decision based on their use of syntax, grammar, tone, and word choice was rolled out. But again, they did not give examples or explain this. 

As someone who admins and mods several creative spaces, this kind of behaviour worries me. Actually,  it terrifies me. Because anyone can be next: you, me, the brand new writer eager to share their work,  or the fandom oldie. 

It sets a dangerous precedent for fan spaces and the policing of creative works, because it leaves writers in these spaces open to harassment, bullying, intimidation, and censorship under the guise of keeping fan spaces "AI Free". Anyone can accuse  someone of using AI to write, and this can be used to bully people out of fan spaces. 

Moderators are supposed to keep fan spaces safe for members, not take it upon themselves to play detective and police creative works in case someone is using AI. In fact, mods taking it upon themselves to police writers in fan spaces by handing out warnings and bans are making fan spaces unsafe and encouraging social exclusion. This kind of behaviour will actively scare and intimidate writers into simply walking away - because who wants to hear an accusation that their work is AI?

Facilitating AI witch hunts is killing fan spaces.

Attempting to police the writing of others is alienating writers. 

The server I run 100% has a no AI rule. That AI rule, however, can only be enforced if someone says they are using it. We don't accept anonymous complaints as proof and don't proceed with issuing strikes because someone assumed. We have open discussions surrounding AI with those who have either a) admitted to using or b) showed interest in using AI. Our rules on it are reiterated. In a couple of instances it led to the parting of ways, in others it's led to a greater understanding.

Because of the field I am in, I have had to learn a lot about AI over the past couple of years. Taken courses on it. Learned what to look for, what not to look for. And I can tell you, as someone who knows AI writing well, there are not always tells.

As an example, recently a friend on another site posted about how if em dashes are used, you know it's AI, because AI favours em dashes. While the point about it favouring them is absolutely true, it fails to recognize that AI favours them because it's been trained off writers—writers who, like myself, employ the use of em dashes frequently and in the correct manner.

And I want to be very clear. I am not in favour of generative AI being used for writing and artwork, creating music and video and imagery. I am fully against it.

But as VorchaGirl highlighted, we cross a very precarious line when we take that strong stance into policing and gatekeeping.

The friend she mentioned who recently went through this in a group is a friend of mine as well. An incredible writer whose work has grown naturally over a year of intense writing. Of not only writing every day, but actively dedicating herself to improving their skills, working with editors and beta writers to help identify areas that don't flow well or where they might be able to improve. I know this firsthand as I have had the pleasure of looking over their work and of providing feedback. They've always been an amazing storyteller, and seeing their writing grow has been a joy.

This witch-hunt for AI writing is getting out of hand. It's not just in the groups. We have seen it on Tumblr and A03 as well. Ironically many of these AI hunters are using AI to detect AI. Something that is not only unreliable but is absolutely abhorrent as they are providing AI tools additional training and content of other writers.

This kind of gatekeeping does more than harm writers in the community. It discourages them from writing. It will chase away people who are new to writing, whose work may mimic some of the less undetectable AI works because they are learning still.

I am livid about what my friend went through. And it highlights a problem on a greater scale. This gatekeeping and hierarchy of feeling like you are somehow better than someone else because their writing doesn't align with what you think is right? It's bullshit. I have my theories about what is driving it, but honestly those distract from the point.

Warning!

It's the sinday after Valentine's and he gives you this look.

'Our Final Dance': A Dark Solavellan Cameo