Since some governments have plans to enforce labels for AI-generated content (text/image/video/audio) on social platforms, should we consider server wide rules for AI generated content (in non-AI communities)?
Or should AI generated content rather be only posted in AI communities?
The thing is, AI Image generation becomes more and more realistic, so I think at some point people can't be stopped to post it in other communities as well. Also for having efficient filtering on it later, it might be useful to have a clear AI-Label (e.g. [AI] in the title) in place.
I feel like restricting AI-generated posts to AI-labelled communities is the way to go.
I'm not a fan of AI artwork in general, but I feel so long as it's being restricted to certain communities and not left to move into other ones, then things are fine there. My main concern with AI artwork is that art is being sampled without the original artists' explicit permission a lot of the time, and while I understand people's trouble with niche interests at times, I feel that that personally curated content from an AI prompt is best kept privately than shared publicly, potentially in some way infringing on the original copyright.
At the very least I would like to see [AI] required in all AI content posts, including posts in AI focused communities. This is because there is no ability to filter AI out and it is not obvious which communities are AI-focused.
I agree with the idea of mandatory tagging of all AI content even in AI only communities, seems like the best way to ensure people immediately know as well as have the option to filter it out (not sure if that's a possibility yet with tools though).
Also it would be good to have that as an instance-wide rule too imo so that people who seem to make a habit of not labeling AI content can face consequences and be held accountable, in order to make potential trickery/scams etc harder to pull, because there's certainly going to be people who will try to somehow profit by trying to pass off AI-generated content as real, as is already happening on many other platforms (and I don't just mean financial profit, because it can be a type of more indirect profit to gather followers and engagement on that basis).
I am not a moderator myself (yet), just a content creator/poster that feels strongly about this topic, so I hope it's okay for me to chime in with my perspective since I think this is a valuable conversation to be had.
It would also really suck for OC creators to be discouraged from posting because unlabeled AI content may be getting more engagement than their real posts, so that's also something that needs to be considered.
honestly, I'm against regulating this instance-wide. Let's try and not become the minefield that Reddit was. Dozens over dozens of (sometimes hard to understand) rules in every subreddit. Adding lots and lots of serverwide rules will only add a layer of frustration to folks who just did this cool thing and want to share it with a community.
I don't see "big scary AI" as the internet-flooding issue it's made out to be and looking around this instance proves me right on that. Why overcomplicate things that aren't an issue in the first place? We can always change this if it becomes an issue.
Apart from a couple of comms I inherited where the "No AI" clause had been in place (such as [email protected]), or the studios-specific comms, I've intentionally left the comms open to the idea of some AI gen content. if the AI content becomes too overwhelming, then that's a bridge to cross later.
I think tagging such content is a reasonable step, so that there is no potential confusion or doubt for the viewers or mods.
BTW, this is going from a guy who has blocked all the AI-only comms as I don't actually have much interest in the current AI hype and fad. I simply prefer to see real human beings as unique individuals in all their glory and shortcomings...
I was against AI images before, but reconsidered. I'm into fur fetish, and on Instagram there are two really interesting AI based sources: One is doublekcomics, who AI-enhances their own (old) comics bringing their creations into the photorealistic world, the other one is (I forgot the name) which shows everyday pictures of men in fur - something where you don't see a lot of content of, outside of the gay fur porn sites. Both are new and interesting.
The main point is probably "are the creators who are using AI as a tool to create interesting works"? Or is AI being used to do the same old stuff over and over again?
However, labelling AI with [AI] is a good idea in any case and it might well be a site-wide rule for me.