The main impact we artists see is that the datasets are built up of art scraped from the internet, without consent, without credit and without any form of compensation for the original artists.
And we see a flood of AI stuff in a volume we can't follow. Essentially making our art less visible,stunting our growth in te process.
AI the way its being used now is a blight on artist communities.
This is not a tool... This is a replacement for artists, on the backs of artists.
I will not use generative AI, I will not support it, I will not endorse it and I will resist it with all my abilities.
As long as there is no ethical way the datasets are built up Generative AI will be my enemy.
I still like quality that I get without having to worry about AI. I know there are those of you who want to try to use it. I myself will probably never fool with it because it runs off the backs of hardworking creative artist it does not think for itself. But to each their own.
My take on the AI craze, from someone who has played with it a bit and have it installed on a Windows server in my local network.
It's a tool.
This is both good and bad. You can have no visual / graphic artist ability but if you have a modicum of ability to write a prompt you can create 100's of images and then cherry pick the best of them. Then you tell yourself, 'I AM an artist!'
You can have a lot of artistic ability and no idea how to write a prompt and create 100's of images to cherry pick the best ones and then tell yourself, 'I AM an artist!'
The other thing about AI generated art, is in the fact that it will only give you the styles of images it knows about. It will not let you 'create' something entirely new, it is always derivative of the previous work of others, so really is the prompt writer the artist or are all those other samples from previous creators used in the dataset the real artist.
Now there is also the same concern that
Willibrordus brings up. In just a few hours, I can create a 100 images in Stable Diffusion, in DAZ and a couple of hours, I might get two images, usually just one. There is no way to keep up with the volume. Also the large quantity that is produced diminishes it value to anyone. So the images while it might be interesting, has almost no value (monetarily). Where I think the value lays is in the dataset and the prompt writer not the 'art' itself.
Now all this is from a person who endorses new technologies, and especially loves to see bespoke creations that help others react. In that sense, maybe AI generated art has accomplished it goal.