Forum
>
General / Nonfiction
>
Thread
Log in |
Register
Well the technology right now is not ready yet and it messes up a lot. but is this something that will be much better in the future or will it only get minor improvements.
like Google translate for example. it got better but it still can't touch someone who actually understands a language.
but when it comes to art that is not trying to be realistic. it gets a lot closer to a real artist. it can do certain styles well that are hard to do for a human without spending a lot of time. but i noticed that human artists have more creativity while the ai will always draw similar things
And i don't really know if i like or hate it. part of me hates it because it's uncanny and it messes up often.
but it also gives people like me with no talent and limited amounts of money a chance to make their fantasies into actual pictures.
...you want AI pictures to be allowed on the main site or am I reading too deep into this?
...you want AI pictures to be allowed on the main site or am I reading too deep into this?
No, I think that would go against the purpose of this site. I'm just curious about what the community thinks about the implications of the rise of AI generated "photos". How would you allow or disallow it, if it becomes almost impossible to discern real from fake photos and what would that do to the site and community?
Well the technology right now is not ready yet and it messes up a lot. but is this something that will be much better in the future or will it only get minor improvements.
like Google translate for example. it got better but it still can't touch someone who actually understands a language.
but when it comes to art that is not trying to be realistic. it gets a lot closer to a real artist. it can do certain styles well that are hard to do for a human without spending a lot of time. but i noticed that human artists have more creativity while the ai will always draw similar things
And i don't really know if i like or hate it. part of me hates it because it's uncanny and it messes up often.
but it also gives people like me with no talent and limited amounts of money a chance to make their fantasies into actual pictures.
Yeah, it's possible that it never will be good enough to be 100% photo real. I'm also not really sure what to think about it. I'm both curious about how far we can take the technology and what it enables, but at the same time it's kinda scary.
An ongoing thread experimenting with the technology, which in itself is an example of how quickly it has improved: https://www.girlswithmuscle.com/forum/thread/...
The improvement of text-to-image AI technology has accelerated severely the past 2-3 months with DALL-E, Stable Diffusion and Midjourney. In my opinion this is heading towards a future where you can just type out a description of what kind of image you want and it will be generated for you.
What do you think this means for GWM? Does it matter that the person in a picture is real or not? Why and how much? How will we know the difference between a real person and an AI generated person?
If you look at it from an achievement perspective it of course matters, as in if you think that feats of strength or taking the human body the the limit is important it will of course matter. But if you are just in it for the aesthetics, it shouldn't matter as much?
What do you think? Does it matter to you? Will this be a problem for the community?
The attached picture is an example generated by Stable Diffusion.
What terrifies me about AI is its capacity and its ability to be used for evil. For example, even with our limited technology, it is more easy to make someone look guilty of crimes they haven't committed with editing software. Can you imagine how much easier it shall be in about twenty years?
No one will ever able to know what the truth is anymore because lies shall become so convincing, even what you see before you can lie to you more deceitfully than a gold digging whore.
But, ooooo, look how easy it is to draw pretty pictures without any talent or ability so, who the Hell cares about all that silly stuff, am I right?
I find some AI images very attractive. We’ve always enjoyed the female form in art historically, why not muscular women? Totally agree with your point about technology being used for nefarious purposes. And you’re right, the scariest part will be that the technology will be used by those entrusted with upholding to persecute and prosecute, whoever the authorities may target.
I find some AI images very attractive. We’ve always enjoyed the female form in art historically, why not muscular women? Totally agree with your point about technology being used for nefarious purposes. And you’re right, the scariest part will be that the technology will be used by those entrusted with upholding to persecute and prosecute, whoever the authorities may target.
Yeah, I've found some pretty nice-looking AI-generated images, too. But, I'm not impressed with the artists who made them. If all it takes is just a few clicks of a button here and a button there and, presto, a full-fledged masterpiece pops out without any effort, well. I can appreciate the resultant image but, the artist is not anyone I'm going to be wowed all that much by.
Like George Jetson at his job, all he really does is push a button at regular intervals all day and, the lightyears-advanced technology takes care of the rest. The technology driven by the button is impressive but George Jetson? The guy is a Don Nadie (spanish for 'lord nobody'). Any jackass can do what he does. So it is with AI-generated art.
New page (not mine), give it a glance --https://www.instagram.com/abacngtr/ Creator is on here, name's abac.
I've looked at a lot of AI art, and written my opinion in a DA journal. Basically, AT THE CURRENT STATE OF TECHNOLOGY (important caveat!), with all the ethical issues on training the AI on real artists who haven't consented and aren't paid (Greg Rutkowski must curse the day SD came out!), and looking at the results, it sucks. The problem right now is we have too many lazy users, I will not call them "artists", all going to the same SD sites, using the same references, and producing walls of identical-looking images. I called it "pornography", meaning "erotic images made with no thought to communicate anything but 'You might like to look at this". More explicitly, it's hackwork and at the present level of AI, I don't see it as anything more than a toy.
Dude, chill out. Remember the website you’re on. This is all just fantasy to begin with.
If a user could give more detailed feedback and editing, rather than just enter a short prompt and hope to luck and the AI, I would think better of it.
I've looked at a lot of AI art, and written my opinion in a DA journal. Basically, AT THE CURRENT STATE OF TECHNOLOGY (important caveat!), with all the ethical issues on training the AI on real artists who haven't consented and aren't paid (Greg Rutkowski must curse the day SD came out!), and looking at the results, it sucks. The problem right now is we have too many lazy users, I will not call them "artists", all going to the same SD sites, using the same references, and producing walls of identical-looking images. I called it "pornography", meaning "erotic images made with no thought to communicate anything but 'You might like to look at this". More explicitly, it's hackwork and at the present level of AI, I don't see it as anything more than a toy.
As a current producer/publisher and former content creator, I get it. Artists want to be paid for their work. Someone used one of the programs recently to create a 2500-word "story" within 2 minutes. He then got the same software to write a "song" in another minute or so.
But as I say with farmers, miners, entry-level coders, truckers, rail engineers, stenographers, factory workers, etc., no job has a right to exist. Some sort of technology eventually comes along to replace us all, so we have to find other ways to work, to make/do/create something we can take to the market to earn income.
Art is not immune to that universal reality.
The theft is a problem, but sadly, that genie's not going back into the bottle.
O, I've no hope of turning back the clock. I only hope to make people see that this is not good art, and aesthetically unacceptable in its current form.
O, I've no hope of turning back the clock. I only hope to make people see that this is not good art, and aesthetically unacceptable in its current form.
The irony of the first (new) wave AI "artists" complaining about the second (new) wave AI "artists" copying their prompts word-for-word is pretty epic.
When quantity threatens to swamp quality, attention must be called to actual quality. With literal walls of near identical pics, for example in our own AI thread in our forum, attention must be called to individually crafted, honest to Odin art. The alternative is Warhol's tomato soup cans.
I still don't get the Warhol thing. I remember thinking for a brief moment during my life that his work was "special", but really, it was not and never was. I just saw a small exhibit of his works (including Marilyn, Campbell's soup cans, Mao) at the Louisiana Museum of Modern Art in Humlebæk last week and was completely underwhelmed...again.
You make an excellent point about the sameness of the things, but how much of that lack of uniqueness is actually problem? The beautiful ones will normally - not always - separate themselves in some way. All other things being equal - as in, EXACT matches - the one will have a "prettier" facial structure than another. Or maybe she'll just have better marketing or be featured on an already-larger-than-most Instagram platform.
As a lover of both art and technology - as well as the concept of paying artists for their creations - this is all tough to reconcile.
My reference to Warhol was in regard to his artistic statement that a beautiful image repeated enough, such as his silkscreen of Marilyn Monroe, becomes hideous. The sensibility revolts.
In the same way, a bunch of nearly identical pictures of beautiful muscular women, our ideal, becomes ugly. As I said, too many users, using the same websites trained on the same artists, produce a sense of revulsion, at least in me. Adding in the lack of effort and the ethical problem, and I can't help but turn away.
O, I've no hope of turning back the clock. I only hope to make people see that this is not good art, and aesthetically unacceptable in its current form.
Any schmoe can enter a few prompts, name drop some artists in an interface, click some buttons and call it a day with the computer doing all the mental and physical heavy-lifting. Not everyone can actually sit down and freaking draw something by hand (whether it's a pencil or with a mouse) and create something spectacular.
Some AI-generated art is pretty spectacular but, the artists themselves? They are nobodies. The artistic equivalent of script kiddies (lazy, amateur hackers who use existing code because they don't have the knowledge or the drive to learn how to make their own).
The improvement of text-to-image AI technology has accelerated severely the past 2-3 months with DALL-E, Stable Diffusion and Midjourney. In my opinion this is heading towards a future where you can just type out a description of what kind of image you want and it will be generated for you.
What do you think this means for GWM? Does it matter that the person in a picture is real or not? Why and how much? How will we know the difference between a real person and an AI generated person?
If you look at it from an achievement perspective it of course matters, as in if you think that feats of strength or taking the human body the the limit is important it will of course matter. But if you are just in it for the aesthetics, it shouldn't matter as much?
What do you think? Does it matter to you? Will this be a problem for the community?
The attached picture is an example generated by Stable Diffusion.