AI art isn’t competent, anyone with moderate education in art can see this, there’s a lot of copying, and not good understanding, it’s not like a collage, it’s more like a remix but the elements don’t marry or contrast properly, it’s big failure is that it can’t do what humans do, in writing it’s said ‘writing is rewriting’ and writers talk about the first draft being the most important step and the least important version of the work. because you can’t edit nothing, and the first draft only exists to be torn apart and edited to death. This is similar to visual art. but we don’t talk in drafts but in passes. you do a pass for line, for color, for composition, for lighting, for balance, for proportions, not in a strict order or anything, and many artists even untrained artists go through these passes without consciously knowing that’s what they’re doing in every case, but it’s similar to writing, the creative process is about changing a work until it is complete, and that requires a lot of time and thought and intent, AI only takes finished images and remixes them. which is why if you ask it to ape cubism or optical art or any of the forms of art that deconstruct elements of art it can’t do it. it takes the image of those finished pieces and it slaps them on a object your requesting. you want a cubist keytar painting in hot pink? your going to get an image of a realistic keytar with cubist finished artwork painted on top of it, not a cubist painting of a keytar, because deconstruction and intent of that level requires human intellect and understanding of elements, and remixing can’t get you there at all. this isn’t magic, anyone could figure out how to make original art in a cubist style or optional style or dada ist or surrealist if they spent a little time studying art elements and the said styles, it just requires understanding that realism isn’t the point of art, and that elements like line, tone, and color can be separated and made to describe the visible world in a way the visible world doesn’t exist as, but AI can’t grasp elements or deconstruction, or how to make a finished piece of art from base elements, it only sees a finished work and copies the appearance.
AI art isn’t competent, anyone with moderate education in art can see this, there’s a lot of copying, and not good understanding, it’s not like a collage, it’s more like a remix but the elements don’t marry or contrast properly, it’s big failure is that it can’t do what humans do, in writing it’s said ‘writing is rewriting’ and writers talk about the first draft being the most important step and the least important version of the work. because you can’t edit nothing, and the first draft only exists to be torn apart and edited to death. This is similar to visual art. but we don’t talk in drafts but in passes. you do a pass for line, for color, for composition, for lighting, for balance, for proportions, not in a strict order or anything, and many artists even untrained artists go through these passes without consciously knowing that’s what they’re doing in every case, but it’s similar to writing, the creative process is about changing a work until it is complete, and that requires a lot of time and thought and intent, AI only takes finished images and remixes them. which is why if you ask it to ape cubism or optical art or any of the forms of art that deconstruct elements of art it can’t do it. it takes the image of those finished pieces and it slaps them on a object your requesting. you want a cubist keytar painting in hot pink? your going to get an image of a realistic keytar with cubist finished artwork painted on top of it, not a cubist painting of a keytar, because deconstruction and intent of that level requires human intellect and understanding of elements, and remixing can’t get you there at all. this isn’t magic, anyone could figure out how to make original art in a cubist style or optional style or dada ist or surrealist if they spent a little time studying art elements and the said styles, it just requires understanding that realism isn’t the point of art, and that elements like line, tone, and color can be separated and made to describe the visible world in a way the visible world doesn’t exist as, but AI can’t grasp elements or deconstruction, or how to make a finished piece of art from base elements, it only sees a finished work and copies the appearance.