Raw LLM Responses

The other thing that is missing from these discussions is, what is the actual future of the use of AI going to look like? To approach this question, I first would like to ask, have you spent any time trying to use AI to create an image? I get the sense that most people commenting on this and criticizing AI art have not actually tried using it. The reason I say this, is that if you have used AI to create content, you would quickly realize how limited it is. For example, I thought it would be fun to use AI to come up with a logo for my musical practice. As a saxophone player, I put in lots of prompts trying to get cool images of saxohpones. Well, you can certainly get a lot of images that have saxophone-like content. Some of them are cool and inspiring. But are any of them even close to being useful as a logo? Not even close. The real role of AI art in the moment is limited to coming up with draft ideas. If you are interested in obtaining a unique piece of art that serves some useful purpose, you will still need to hire an artist. Except, instead of describing what you want in words, or showing examples of art that is similar to what you want, you may be able to show them a set of AI images that each demonstrates some useful elements that you like. There is no way currently of telling AI to synthesize concepts among a set of its output. You have to go back the original prompts and tweak them, and you will get a new set of images. Current AI models cannot simply retain some aspects of previously generated images, and modify only peripheral elements. There are 2 main ways in which the current state of AI art is lacking. One is the quality--the actual images tend to be small and look terrible if you attempt to blow them up, as they are not production quality. Another is detail--while some of what AI produces is remarkable, the more you look at it, there is usually more that you want to "fix" that makes the image unsuitable for the intended purpose. Of course, part of what we are doing here is thinking about the future, and so some of these issues related to AI utility will start to be addressed. In particular, in terms of image quality, commercial AI products will not have difficulty producing things of the size and depth needed for commercial purposes. But I think artists should start thinking about how learning to use AI image generation will enhance your skills as an artist, and improve your workflow. Instead of having do do actual drawing to generate ideas, as a super-user of AI content generation, you will be able to quickly help your clients identify what they like and don't like, and be able to hone in on a concept that is worth speniding your time and skills, and their money, on. So, what I have described here is more along the lines of artists who are commercial artists or illustrators. For the cases of artists who are making fine art that will hang on a wall, then you are not even competing with AI yet, as it is not yet in the business of non-digital assets. But when that time comes, since of course it is possible to print onto canvas, the same issues remain--will you really want to print out an AI piece that has weird eyeballs and hands with 3 fingers, or will you want to bring it to a digital artist to treat it as a draft and create a curated image that is worth memorializing on a wall?