Alright, I’m back, and I think I’ve discovered what was bothering me.
The points of yours that I now agree with, at least to an extent:
That AI doesn’t technically take actual identical aspects, just “learns” from existing artwork.
That it is fiendishly difficult to change the laws surrounding copyright so that AI art is banned, and fanworks are not.
I have three points in rebuttal, primarily:
Laws are not necessarily promoting the most ideal behavior
When I write an essay using sources, I cite them. Not citing the sources I learned my material from is considered academic dishonesty, and I am not allowed to do that. When you look at a scientific study, even, to cite an example not implicitly within the realm of “teachers want to make sure your sources are valid,” they cite their sources. And what they come back with is a product built upon others’ work completely ethically.
AI does not cite its sources. It does not communicate to the viewer where it obtained its information. And the reason that this differs from human-created artwork, where people don’t cite their sources either, is that the AI is not thinking for itself. It does not produce any original material. There is no individual touch. It spits out data in a pattern it has been told to follow. And you may say that makes AI artwork more of a tool or a style than an artist, but I would also like to argue that there isn’t nearly as much of a hand involved in the work as there is in, say, photography or Impressionism, even though those are both famously art forms derided when introduced because of the lack of work critics deemed they involved.
Laws are fiendishly difficult in any situation. They have to say “this behavior should be followed,” “this behavior should not,” and do all of that in a manner that does not allow for loopholes, unintentional interpretations, or future laws that allow for something the lawmaker does not want. Laws are not the ultimate decider of morality. They never have been. Just because something is legal, doesn’t mean everybody should or will like it or agree with it, or the opposite.
I am going to give an extreme, black-and-white example, that I do not believe is in the vein of AI artwork in the slightest, to demonstrate that point. I repeat, NOT comparing this to AI art, simply showing how laws do not and should not equate to morality.
Nazism was a political party. All of its actions were legal, at least in Germany at the time. Jews could be legally thrown out of their homes and property and sent to concentration camps. Those objectively evil actions were legal, and were actually illegal to circumvent.
Laws are NOT equal to morality, and they never have been. Hopefully they trend in that direction, but the primary point of laws is to denote what behavior is and is not viable for a successful society, at least from the point of view of the lawmakers.
And, furthermore, people on the internet having a view of something does not inform legalities. I can believe that lobbying is bad because it means that those with more wealth get to determine laws; that does not make it illegal, or mean that a lawmaker is going to introduce a law that makes lobbying illegal. Someone else could believe that taxes are theft; that does not mean the government is going to eliminate taxes.
Lastly, recognition is relevant here. When I buy a bookcase from Target, I know it’s not equivalent to the work a carpenter would put into it. It’s made of dubious materials, thin sheets of cardboard backing, and a few nails and screws that I insert myself. I can very obviously tell the difference.
Same with clothes made from sweatshops, and self-checkout. None of those are the same experience as an actual human putting in the work to create the thing.
[And, furthermore, sometimes to exist is to participate in unethical systems, even when I try to avoid them to the best of my ability. (Was I supposed to tell my first job at 16 not to mass-order camp T-Shirts from a sweatshop? Am I not supposed to use a self-checkout machine when I need something cheap and sensory-friendly to eat and I’m at the airport inside a completely autonomous shop?)]
But human-created furniture and shirts, and human-ran grocery stores, still exist. And I can find them, if I believe in the value of those things (and can afford them, in a society that constantly wants me to pay more for less when it comes to things I need to survive). The problem with AI is that it is nearly indistinguishable from human creation in many cases, because of how it learns from an inherently human product.
I am able to remove certain things from my experience completely, all by myself. I can pick a cash register with a human, I can buy a real wooden desk, I can buy ethical clothing. It is impossible to ensure that the art or book I am purchasing was not created with AI. There are certain tells, but those will be going away when the technology improves.
With Impressionism and photography, you can tell that the product is different. I can tell Tumblr to hide any posts which mention the DSMP from my homepage. If people choose to pass off their AI as something which they made, I cannot choose to disengage from it. And when somebody sells an AI-generated book on mushroom foraging on Amazon that gives literally deadly advice, people will buy it, and people will be killed by it.
When people use AI, they are training AI. And the more AI is trained, the sooner we will be entirely unable to extricate ourselves from it.
That’s one of my biggest problems with the usage of AI.
(I do, however, believe that harassing well-meaning people is bad. As is any harassment towards anyone that isn’t “I disagree with you about this thing and I am going to say so.”)