Let me put a straw man, and try to find a middle point, when the copyright argument stops being applicable:
1. A painting was done by an artist.
2. On a computer.
3. With a help from an image processor software.
4. Using some advanced filters, like super-resolution, that utilize computer vision techniques. Like neural networks.
Many smartphones already automatically process your* photos with some advanced CV algorithms. That can be called "machine generated art".
I'd personally prefer to stop saying "neural network did X", same way as we don't say "a bulldozer built a road, a crane built a house".
Even non-generative-AI inside Photoshop only mutates images. Generative AI is the source of images.
Or, is the distinction you are making based on there being an image before the model is used?
Let me put a straw man, and try to find a middle point, when the copyright argument stops being applicable:
1. A painting was done by an artist.
2. On a computer.
3. With a help from an image processor software.
4. Using some advanced filters, like super-resolution, that utilize computer vision techniques. Like neural networks.
Many smartphones already automatically process your* photos with some advanced CV algorithms. That can be called "machine generated art".
I'd personally prefer to stop saying "neural network did X", same way as we don't say "a bulldozer built a road, a crane built a house".