Raw LLM Responses

The actual issue I take with AI has nothing to do with the "ethics" involved. There's already plenty to be said about art theft and legally it's already such a moral quagmire that I just don't feel it's worthwhile to try and argue with the type of person who will scrape an entire database of images to train their model but then throw a fit because someone "stole their prompt." As far as I'm concerned, the modern landscape is a shade of grey so dark it might as well be black. AI art is theft in 99.9 percent of cases. I really wish these models would provide a data sheet that spits out what images and artists were used in the creation of an image so that, not only can these artists get credit, but also allow people to FIND these artists and raise their profile accordingly. Like imagine an alternative that paid a royalty or commission, even something as little as $0.30 cents, split between every "source" for each generated image. All of sudden those millions of images being generated would accumulate a fair chunk of change and exposure for artists whose art is being used in these models. For me, it's that AI generated content becomes a crutch for people who obviously have some interest in the medium but don't want to put in the work to improve. There's always going to be a massive failure rate of hopefuls in any creative field, but I worry that by creating this path of least resistance it has inadvertently begun to divert people who might potentially become successful artists before they even get to step 1. It's salting the earth so the next generation of creatives cannot grow. That's not to say I don't think the Anti-AI crowd are entirely in the right though. AI is the future and will be a massive force in many fields, it's simply being used and abused by people who don't understand its use cases. AI DOES have a place. It's as a support to existing processes so that artists can work faster and more efficiently using skills they've already built up. There are some projects that are actually quite incredible for doing the mind numbing busy work specifically in the field of animation (automatically filling in base colors, inbetweens, compositing, etc). Also, the term AI is often being misused or misunderstood when "AI" is being applied to extremely different technologies that already existed prior to the advent of ChatGPT and other AI models. So now old technologies are getting attacked simply for adopting the label. For instance, I find the backlash against artists who get "caught using AI prompts" is a step too far in the opposite direction. The arguments against it rub me as a bit pedantic and luddite. To me, it seems entirely reasonable to allow AI to provide several varied compositions as a base for someone to then do themselves but with a steadier, more skillful hand. So long as that artist has already built up their skill set to polish it into a better finished product, effectively all the AI has done is helped them visualize something they were already capable of making but faster. It's only a problem when that artist begins to solely rely on it for composition/coloring/line art/sketching/etc (whatever step of the process they rely on it for) as a crutch in the same way I mentioned for fledgling artists. It's still vitally important that they maintain their skillset so that the AI has something with which to provide a more diverse output. It reminds me of the type of person from the mid-2000's who would accuse "tweening" (automated "in-betweens") in Flash animations as "not real animation." Sure, it's a shortcut, but often artists would use that technology to incredible effect using rigged models that eventually became an industry standard in televised animation because it filled this perfect niche between quality and efficiency in the animation workflow. We need to figure out where that happy middle ground is for AI and it's simply too young as a technology for us to know for certain where that is.