Since Lemmy is still in it's social infancy, it may be a good idea to handle AI generated content appropriately
Since Lemmy is still in it's social infancy, it may be a good idea to handle AI generated content appropriately
I'm pretty neutral on the topic of AI generated content but this is an extremely hot topic for many creators right now, would it be possible to create a filter system to disable AI-Generated Content? For starters it could be manually flagged or communities could be fully flagged?
You're viewing a single thread.
Is that really necessary outside of art communities? That seems like something they could just make a rule about and enforce on their own.
Like if someone generated a picture, and then made a meme out of that picture in a meme community, you'd want that labeled and filtered? Why?
It's a touchy subject for a large contingency of creators and AI rules are enforced by nearly all art communities in there guidelines (if they have them). A rule manually enforced by a large quadrant of communities is a good example of when a feature could be useful.
I'm pretty confident that art communities with AI content rules make up a very small percentage of overall communities, but even if they didn't, it's already either not allowed by them or required to be labeled, so what would this accomplish?
AutomationAssumed it's already resolved rather then manual work (per community). It would also allow people who are adamently against AI to filter AI content in subs that could now allow them, following the filtering ruleset.Edit: Automation was a confusing word when not talking about actual automation
I get fighting against industries trying to replace people with AI, but if someone is that sensitive about AI content that they don't ever want to see it in any context anywhere they should probably try to get over that, it isn't going anywhere. People were mad about sound being added to movies too.
How are you going to automate it without AI?
I don't understand what your referring to. I just suggested we create a filter to hide AI generated media rather than community-lead duplication of rulesets for doing so. In not asking for a neural network that automatically detects AI.
"Automation rather then manual work." How can the filter operate, if "AI generated media" are not flagged manually?
Ok, I see. I used a poor choice of words, my mistake. I meant this in terms of moderation work. Since the platform would provide tools for user to mark their work as AI generated, it would reduce the administration per community. I'm struggling to think of a better word? Assumed Resolved?
I was thinking the moderators are going to have to flag posts as AI-generated , I'm not trusting the people posting them to do it. But 100% agree with what you were saying, there should be a way for users and moderators to mark a post as "containing AI-generated content", similar to NSFW; and then a way for us to filter "AI" like we would NSFW.