3 ms·
It really needs to be said that AI "artists" have confused productivity with quality. I actually don't go to DeviantArt to see your ai generated garbage. I care
by __loam 2y ago
It really needs to be said that AI "artists" have confused productivity with quality. I actually don't go to DeviantArt to see your ai generated garbage. I care more about people who are willing to do interesting things with their medium even if they takes a lot longer.
- jsheard 2y agoYeah, it's almost comical the degree to which quantity has become emphasized over quality. More than a few times I've clicked through to an AI posters profile out of morbid curiosity and seen that they have thousands or even tens of thousands of uploads despite being active for less than a year or so. Even with the supposed productivity boosts that AI brings you can't convince me that someone posting 20+ pieces every single day like clockwork is putting any real consideration into them, but the magic of AI is that something with little time and zero thought put into it can still be superficially passable.
- __loam 2y agoI am a bad artist but I do make art, and have been trying to make art over the past year or so. I've made less than 200 pieces over that time but I can still go back to work that I've spent hours on and remember the decisions I've made and the specific works that have helped me grow or that I'm proud of. Do you think AI artists remember the work they've produced?
- arvinsim 2y agoThey will not remember the prompt that created them but by default AI art encodes that information into the image metadat.
- chefandy 2y agoNot that same thing, though. Reproduction isn't the purpose of considering artistic decisions you've made-- it's to reflect on and refine your eye, ability to communicate things visually, and your trajectory as an artist. That capability is entirely separate from the medium you use, or the source of your work. Indeed, nothing in AI image generation makes that less possible than in physical media, but metadata has a very different use case. The entire point is that you need to pay attention to your output enough to parse the extreme subtleties, and posting dozens of pieces per day negates that.