Raw LLM Responses

I use AI art and intend to expand my use and generation of it. I can't afford commissioned artwork, but I also can't afford paid GenAI so while I support both, there is no money involved. I support legal initiatives which would prevent GenAI from using copyrighted material, or any material not licensed for AI use. I support legal initiatives which prevent companies like Adobe or Pinterest from claiming the right to use your content for AI within their ToS. I don't support people calling themselves artists for only using prompts. I don't support people doing commissioned artwork, then committing fraud by using AI to generate said work. I believe that counterpiracy tools like nightshade are absolutely necessary, even if I believe they will be insufficient on their own. Why do I support AI? The pathway of development for AI image gen seems very clear to me that, IF we can get the proper legal protections for artists, it will spur a new sub industry for artists whom will commission and train their own models and sub-models to sell. For example, one might train a specific sword to reliably appear in a prompt, or a custom outfit. With a greater deal of control over the image, more artistic elements are introduced, and while many of these decisions may be made BEFORE the image is generated, rather than during, the amount of decisions is important as discussed in this video. I don't believe AI generated images will ever be a suitable replacement for professional art in most contexts. However, in some cases, it has very strong use-cases. Take a graphic novel: While its obviously doable for a single person, it has an astronomical time investment associated with it, and depending on the length and style of your novel, most of this time is spent on frames which readers will look at for a couple of seconds. An AI model trained on your own work allows you to generate the images which will take seconds to view in significantly less time. These time saving tools allow you to spend more energy on layout of your page, the writing, and your key frames which require more attention. Is the result going to be as good as a fully handcrafted result? Well, that depends on if you would otherwise be using time saving methods like drawing quick and sloppy, or if time in general is a factor in quality. Is this any different from hiring an art-house whose goal is to faithfully redraw characters based on reference images without any creative control? I would suspect not. A writer is still an artist, even if it isn't visual medium, and so I still see AI generated work which enhances another work to be "art." I'm not going to ooh and aww about someone's ability to retrieve an image from a generator, but I AM going to enjoy a story enhanced by visuals, regardless of how they are made. If an AI assisted process in creating a graphic novel takes 1 year to produce for an individual rather than 5, that artist can create 5x the work. They could create a full series within a decade which otherwise may have taken them a life time.... and sure, teams could do this, but artistic expression has a habit of getting muddied and lost the more people are involved in a piece of work There are plenty of reasons not to use it, but I believe the ethics arguments are weak, so long as we are not committing fraud. Instead we should be focusing on the legal initiatives surrounding it to ensure that AI is only used with licensed material.