furry_irl

1673 readers

190 users here now

"For the fur in u"

Welcome to Furry_irl, a community for furry memes, shitposts, and other relatable images or comics.

Community rules:

Code of Conduct — Follow our instance rules.
Post formatting — All titles should be a single word, followed by _irl. An emoji may substitute the underscore.
Credit artists — If it's not your art, include who made it in the title or the post body. Links are appreciated, except to X/Twitter.
Stay on topic — Images should contain or be related to furries. Images should be relatable or a meme. This isn't the place for general art posts.
Avoid AI images — Our fandom has countless artists, please share their (or your own) labors of love instead.

founded 2 years ago

MODERATORS

liam@pawb.social

Foxfire@pawb.social

Tag_irl (OC) (pawb.social)

submitted 5 days ago by Foxfire@pawb.social to c/furry_irl@pawb.social

20 comments fedilink hide all child comments

Happy 6/21 everyone! Keeping up the tradition of a yearly Esix for the big day.

This is kind borderline for furry_irl, but I think with people memeing about 6/21 a bunch and with how small of a window it is, I'd prefer it here over furry. What are your tag(s)? >:3

you are viewing a single comment's thread
view the rest of the comments

[–] aqua_cat@pawb.social 5 points 5 days ago* (last edited 5 days ago) (2 children)

The problem you encounter is making it money efficiant since e621 is mostly non profitable besides selling the name and shilling the parent company while big AI models are REALLY power hungry. Mi****oft restarted whole nuclear power plant to manage energy demand.
If you propose a smaller model there is a problem of human expercionism. In other words every artist has his own artstyle. So you need the model to recognize EVERY possible artstyle which means you need to train it on EVERY possible artstyle which is not possible right now while making the AI somewhat usable.
Now let's set all previous problems aside and say they have the power, knowlage, time, will and money. You now need to actually train the AI. For that you need the training material. "No problem" you say to yourself. You run the biggest furry art archive on the internet. Well, that is the thing it is an archive, not your art. So you would need to ask EVERY artist whose art EVER got uploaded to the archive if they consent for AI to be trained on art they made, which if you can contact every of them I guarantee 95% will say no. And I belive you now need to do the same for people who taged those art pieces to give the AI something to learn from.
with all that work you are left with mediocre algorithm who is 80% accurate 20% of the time. .
In conclusion AI is not a silver bullet for every problem, not even something like tagging since they are really not as amazing as big corpos make them to be.

TLDR >!AI is not a solution from moral, practical and economy standpoint!< . Of course I am open to other's standpoint so correct me if I am wrong anywhere. :3

[–] piccolo@sh.itjust.works 5 points 5 days ago

while big AI models are REALLY power hungry. Mi****oft restarted whole nuclear power plant to manage energy demand.

Training models is very enegry hungry. But using models can be ran on consumer graphics card. I suppose you can argue to include the training cost to using models, but the models are going to be trained rather you use them or not, so the point is just moot.

[–] CheetahHybrid@pawb.social 2 points 5 days ago

I don't entirely agree, but your points are valid. I in no way believe E621 is non-profitable. If it wasn't making money, it would go away. Second, it should 100% be an opt in system. It should also remain proprietary so nobody could use it outside E621. Getting artist on board wouldn't be too hard if you made it clear that its explicitly for image tagging and not for generating spoofs of their art. Make it up front and clear that this is only for enhancing e621 and not stealing from them. I'll just skip over the training point because there doesn't seem to be a flaw in that logic. It will be expensive to train. I do dispute the accuracy numbers. Since this Ai would only suggest tags, it doesn't need to be 90-100% accurate to be helpful. It just needs to put ideas for tags in the uploaders mind. As more tags are added to images, the Ai could also gradually improve.