How do you all feel about AI generated art? (rule)
So a recent post garnered some comments and reports for being AI art. In light of this we should all have a little conversation about how AI generated images should be handled in the future.
I think we all agree that AI images that are "garbage" or don't add anything should be removed, but clearly some feel very strongly that all AI art should always be removed.
It should be noted that the rules as written and as agreed on by the community does not blanket ban AI, it merely says AI art should be avoided, while many other rules say no this or that instead.
Things to discuss:
Does it matter if an image is AI? Does it always matter?
What about images that are AI generated, but have been modified by a human?
What about images where it's hard to say for certain that it is generated? Me and the other mods did not agree on whether the recent image was AI f.ex which makes it hard to make a decision on whether or not to remove it.
It can be stressful to artists to be accused of having used AI. If we are too militant on weeding out AI art it could be harmful as there will no doubt be some false positives.
Should AI posts require being tagged in the title? (and of course be required to be of a certain level of quality)
I think a lot of us mods feel that AI should be allowed so long as it is not low quality and serves some purpose (being entertaining f.ex), and that the community should not be flooded with AI. What are your thoughts?
Edit: Thank you all for your input! Most of the others are sleeping right now I think, so nothing is likely gonna happen until later today.
You're viewing a single thread.
91 comments
oki, here's my stanc
(gen-picsmeanaiimageshere)
i wud enjoy gen-pic ban
bt its almost impossibl cuz gen-pics r supr gud now n hard to tell if fake
new image gen is not prompt -> result anymor - but mor lik an LLM puts together multipl elements n iterativli improves images, lik specificlli addin text to onli smol bit--- "source" here
this makes it evn hardr to tell if imag is gen pic... which i thinksies is bad
mayb we employ som gen pic detector if we're unsure?seems lik best option, cuz evn i, who kindsa knows how stuffsies wrork,cnt tell anymor ;(
bannin gen pics alsuu removs funi shrek stealin from white house found footag... which... im oki with- bt its kindsa a loss kindsa
so yis, im for gen pic ban n alsuu wan to hav detection tools if were unsur - if u wan ai 196 head ovr to lmmi wrorld i feel...
alsuu - if gen pics wud be fulli allowd - id hav to go outta my way to hide evri singl one when i see one... which i dun lik ;(
a gud way to think bout it is dis:
when i read text n go "oh... dis is llm genratd" i immediatli stop readin
samsies for gen pics
My dyslexia nightmare.
huh? whadds da problm?...
I cannot read your comment without great difficulty. I'm sounding it out like I'm learning to read lmao
I had to use AI to make your post readable
My point is that AI chatbots IMO aren't completely useless, you just have to know what they're good for.
For example, they perform well on translation, many times even better than stuff like Google Translate since they are able to take in contextual cues, correct their results to have better grammar/be more natural, and can adapt based on the user's instructions.
Also, I recognise that LLMs alone aren't really reliable with factual information. That's why I use sites such as Perplexity and Scira, which cite web search results. I also often check those citations myself for further context.
... also, why do you type like that? Is it your way of adding a signature to your posts, or what?
yis its kindsa a signatur thing... many nice blahaj zone peeps are oki with it or evn lik it.. - - -
alsuu lik - i get it.
i personli am kindsa... an llm fan. hate coops pushing it into evrithin, bt i lik em when i need em - so yis, i cn totalli see thad, llms cn be useful.
but lik - dis post is bout image gen.. n u didn say a thing bout the image gens stuffsies besides on ur own commnt... so u sayin "i needed llm to translate" isn funi n alsuu isn lik - addin anythin to the conversation... to me, it jus felt lik an insult ;(
othr question: do u evn interact mch with the 196 on the blahaj zone?... cuz lik - dis post is vrri mch mor orientd at the peeps usin the blahaj 196 thingy here..
the reason im askin is cuz i dun rembr seein u in commnts, nt cuz im lik "u prolli jus here to mess stuff up". if u rlli use 196, thn its totalli fine! ~ <3
Good points Smorty. 👍 I'll try to be more considerate with my communication in the future.
Also, I did see 196 posts quite a bit when scrolling through the front page a few months back, but I wasn't really subscribed to it. And even then, I wasn't a frequent commenter under them. I've just checked Lemmy again after a multi-month break, so I probably didn't see you around.
I personally prioritise clear, grammatical communication, especially if it's not in an "instant-chat" setting like Discord, but if the community you're in is lax about this, it's like, whatever. At least it's understandable enough for the LLMs to glean with 99% accuracy, so... 🤷
Have a great day/night. ☺️
I think you're lying. I think you were able to read it just fine on your own.