Raw LLM Responses

I tend to disagree, AI is great for art, and I say this as a seasoned oil painter. In most cases the generated image is garbage and won't be a final product. But there are some cases for this. Things like stock photos or logos, images for usage in boring business settings, company websites, advertisements and so on. This kind of art is generally trash, and it's a boon for humanity that these tasks can be replaced with the click of a button. For AI as a tool, I think there's a misunderstanding. It's not a tool for creating art, rather, it can become part of the process. For example you can use it to generate a reference photo, similar to what you might do with Google Image search, but in this case you can get a more specific reference image and upscale it to 4K, and change the lighting and angle. Very useful when learning a new form or subject. It can also be used in the process to iterate your own art. Suppose you have an idea drafted at 1:4 scale, or just some napkin sketches. You can feed these hand-drawn ideas into the model and use image-to-image generation to create various alternatives, see your own idea from different angles, different styles, different approaches to lighting, colour, shadow, setting and so on. In that way it can help to bring forth fresh concepts or ideas (for your own work) that you may not have considered.