Same argument could be brought against the printing press. In a sense, it was - that's how copyright became a thing in the first place.
Same for photocopiers, printers, general-purpose computers. If, at any point, some of that tech had "never have been invented or at the very least not released into the public", we wouldn't have e.g. photolitography and thus no modern microchips.
Copyright are a legal equivalent of a dirty hack to preserve some legacy behavior, that became a permanent fixture over time. Like with dirty hacks in code, you're going to get different responses from different people, depending on situation.
The first large scale use of the printing press outside books was to print indulgences:
> In the teaching of the Catholic Church, an indulgence (Latin: indulgentia, from indulgeo, 'permit') is "a way to reduce the amount of punishment one has to undergo for (forgiven) sins".[1] The Catechism of the Catholic Church describes an indulgence as "a remission before God of the temporal punishment due to sins whose guilt has already been forgiven, which the faithful Christian who is duly disposed gains under certain prescribed conditions…"[3]
This directly lead to the reformation and two centuries of religious wars in Europe which proportionally killed more people than both the world wars combined.
The printing press has more blood on it than ink. By contrast the worst you can say about art models is that they are 'derivative' whatever that means.
Indulgences also paid for the art of the renaissance. It was because pope Leo X was so into art that he stepped up indulgences, causing Luther to protest. So renaissance art has more blood..?
i have to ask, do you also count world war victims as victims of the printing press because military orders and propaganda were printed on paper? is the the telephone, copper in the wires and electromagnetism itself also complicit in genocide, because they were used for communication in nazi germany?
> the technologies you mentioned helped spread very useful information, not worthless images devoid of meaning, artistic expression and thought
Both. They did both.
Cameras replaced most portrait painters, but they also gave us "Paint Drying" which is 10 hours 7 minutes long film of exactly what it sounds like — deliberately worthless images devoid of meaning as it's a protest.
And when I was a kid, the talking heads in the press were upset about cable TV introducing the UK to the cartoon channel, with a similar argument to yours.
i'm sorry, but you have to be extremely naive to think that all these companies actually respect people's work enough to verify that all their training data is properly licensed, and deliberately hinder themselves and put themselves behind their competitors by avoiding any copyrighted work
But even if by "theft" you mean "copyright infringement" there is a strong argument for fair use, we humans consume large amounts of copyrighted content and are influenced by it in everything we do, I don't find it unethical or wrong in general, and I don't see how it should be any different for a machine, unless you are against job automation, but in this case you seem to be worried about "theft".
Edit: HN thinks that I'm posting too fast so my reply to the comment below here ahah: "Well, then we agree that it wouldn't be a problem if the model didn't output an image from the training dataset, which is extremely rare or essentially impossible with today's dedup steps used during dataset creation."
models will straight up return training data sometimes if you prompt them with like an image file name like IMG_1234.JPG though
i'd say that this is not the equivalent of a human glancing at an image and then have the memory of details of that image have some small influence on their way of thinking and imagining and creating, i'd say that's theft
also obviously depends on the case, copyright law is often very stupid and broken, like there's no reason why this book written 100 years ago isn't in public domain, but scraping millions of images from artists' websites kinda is
Same for photocopiers, printers, general-purpose computers. If, at any point, some of that tech had "never have been invented or at the very least not released into the public", we wouldn't have e.g. photolitography and thus no modern microchips.
Copyright are a legal equivalent of a dirty hack to preserve some legacy behavior, that became a permanent fixture over time. Like with dirty hacks in code, you're going to get different responses from different people, depending on situation.