General

i was searching for this comment.. the amount of A.I generated text and pics make me wanna puke. id rather someone’s terribly drawn wg inflation art than A.I slop

The difficulty with AI generated stories, is that they are almost impossible to detect with any accuracy in an automated way.

I have an idea though, but am not sure if it would work or not. What if readers could vote whether they though a story was AI generated or not. So say after 10 votes you might get a indicator stating '40% of readers considered it AI' or something similar.

Any opinions on this?

I think this is a good idea to start and then we can refine it once we have an idea of how effective it is.

Would there be any intent to remove it at some consensus point? IE. A 75% threshold over a month of viewership after reaching the minimum vote count?

Just labeling it is ineffective. It needs to eventually also be removed, in our opinion.

We wouldn't remove it, since some members may find it valid, however we'd flag it and make it possible to exclude AI generated content from a search if desired.

This is a problem, just from a technical limitation standpoint. Even avoiding the ethics questions, which are actually quite clear, cut, and dry...

It *will* clog your server storage with a potentially infinite amount of posts. You can generate content fast enough to basically post infinite amounts of pictures and text at such a massive throughput that you *absolutely will* eventually have to do something about storing/hosting it all.

It may not seem like an issue now, but you will get mass-produced garbage posts weighing down your operation and increasing the cost of your operation, all while providing a marginal benefit that is offset by a ton of users, like the people in this thread, finding defending it to be an act of ill will. Even if the AI posters pay, the money in this is finite. You'll lose out in the end to this imbalance.

I'm someone who uses AI and I'm 100% in favour of having a tag, and I think that misleading content, which pretends to be real and does not label itself properly, should be reportable. I do however support keeping AI content on the platform as it can be another tool for creativity in the space. It should be accessible for those who are open to it, and avoidable for those who aren't.

why not just put it in a category?
we have categories for M F T
we are categorized by weight and age and height.

i say keep it but keep it in its own place

I’ve spoken with Hiccupx about this, and I believe we’ll be implementing an option for content owners to tag or mark content as AI generated.

In the pictures section AI content already has its own category. However, the stories section works slightly differently, so using a tag system would make more sense.

This will, of course, rely on uploaders being honest about their use of AI. However, having content clearly marked will allow readers to filter AI content either in or out, depending on their preferences.

c00kie
FF Team