As someone on another forum an artist themselves pointed that being an artist isn't the most stable occupation to begin with,
Van Gogh would agree, despite nobody datascraping his art while he was alive.
Generative AI has allowed the masses to do something only a few were able to do before, is this a good or bad thing? Well that depends on your world view.
From a utilitarian point of view, it would depend on the balance between this good (everyone gets access to something they couldn't) and the consequences. Much like the classical argument against yelling "fire" in a theater, freedom of speech is great and restricting him is a little detriment to every theater-goer, but the few persons who die stampeded following a false alarm are subject to a detrimental effect, and we need to balance the two. Lawmakers are just trying to determine what the public opinion is over this balance, so in some countries they will say "Free speech is paramount, a few people stampeded to death don't warrant restricting it" while in some other country, or at other time, they will determine that "Getting stampeded to death is not nice, let's ban yelling fire in theater, because no-one can say they are really harmed by being forbidden to yell fire when there is, in fact, no fire, and they happen to be in a theater". I don't think the legal determinations make anyone morally wrong, irrespective of the outcome selected. Same with the debate in this thread: there is a balance to be found between the benefits of datascrapping for all (the people enjoying free speech) and the detriment it inflicts to artists (the stampeded theater-goers), and the calculation of this balance will be different in different places and time, taking the overall situation into account, and I don't see any result being morally bankrupt.
Folks spend time learning their craft/hobby be it from watching Bob Ross on youtube or taking a class at night or just learning as they go but along comes Fred who used Dall-E 3 along with hours of prompt tweaking, in-painting and generating to get his prefect image should Fred be told to "bugger off" just because he did something differently? I see AI as yet another means to an end, which is allowing people to do something that was once limited. See the thread over in the D&D section on Dall-E 3 and my own prompt thread here.
May I disagree? I wouldn't think they get the same result. Or it would be saying that anyone can be much better than Usain Bolt because they can drive a car much quicker. A bespoke suit hand-made by a tailor might not be fitting better than a cheap machine-made suit (well, often it will, but it may not), and they serve the same end, but people might assign a different value to the bespoke and the ready-to-wear suits. Handmade is a distinct quality. They fill the same need (being protected from the cold, being recruited in a bank or other office where arriving wearing a t-shirt will be frowned upon...) but they are distinctly different items. I can see it with artpieces, where value can be assigned to a real painting, where a different value would be assigned to AI-produced art that would be printed and framed.
Companies and consumers will always chase the cheapest way to do something in order to keep that bottom line/spend the least amount of money.
That's generally the case. However, I remind you of the Veblen effect: Veblen goods are (generally luxury) goods where demand increase as the price increases. I think the same thing might happen with art. Or is already happening with art. I might have not paid the same price for my lithography of a real work of art by a real artist than I would have for an AI-generated image. Yet, it's just a lithography, not something that was hand-drawn. The market would value it a lot more if it were the original drawing.
People in third world countries have always been exploited for cheap labor so them being exploited for data training isn't anything new other than the job. There's no other way to say it without sounding like a naughty word.
Maybe the error comes from considering that producing art is a commercial endeavour. Most of the art I can name wasn't protected by IP laws (not invented yet), and wasn't created with the intent of selling it for a living. Also, I have a friend who paints, he's doing this for his enjoyment, and not with the intent of making money. Yet I value a lot the painting he presented me with. I don't think we should reduce art-making to the "commercial art" segment of the art production.
The laws have always been slow to catch up the pace of technology so nothing new there. The internet was around since the 70s but it wasn't till 98 that DMCA was created and it wasn't until 2000 that it became the law we still debate about it. And we are still trying to make laws regarding the internet and content. I'm all for laws that are fair to both creators and folks who using AI to do what they wish to do (with in reason). But I don't have an answer to how do that, this is also my answer to the issue of data scrapping and how to make it fair for everyone..
The key is that there might not be a single answer. You mentionned the case where market forces should be used to promote an ethical stand, where people shouldn't buy products that are underpaying people. But what is underpaying? Is a German factory worker underpaid, compared to its Luxemburgian neighbour? (Luxembourg has, as far as I know, the best minimum wage in the world, taking purchasing power parity into account). Should German not buy German car? Should they stop owning a car, because to be honest, Luxembourg doesn't make a lot of cars. Should Swiss not buy German cars? And there is the charity aspect of it. If a Swiss buys an American car, he helps American factory workers not to fall into dire poverty, so there is a value in his militant buying of a product. Same with a Chinese car: sure, the employment laws are worse in China, but for some agricultural workers it is an improvement to work in a car factory rather than half-starve on their agricultural lot. I think purchasing behaviour can't be simplified to a binary "do we support the workers or not?"
When there's a tech upheaval there is also a disturbance in the job market, only this time it's sudden like an earthquake instead of a slow roll like coal miners and green tech, which begs the question of: Why should artist be afford protections that other jobs weren't when those workers lost those jobs due to technological progress?
Especially in a context where other jobs are as threatend by AI as artists. Uber drivers might not resist the advent of the autonomous car -- and if some may think we'll never achieve safe autonomous car, let's speak about rail conductors, where the trains are moving on high-speed railways never crossing anyone (or where crossing is already computer-controlled). But more specialized jobs also are threatened. You study 10+ years to learn surgery and all of a sudden, robots may operate patients better. You won't get any special protection. If cheap drones can become intelligent enough to move crates, say goodbye to delivery men and warehouse workers, and so on. They won't get any legal protection, unless we imagine that the collective choice in a country would be to ban any form of artificial intelligence, like the Dune world banned computers.
Instead, we must ensure that displaced workers have a safe place to land. That their skills can be redirected into adjacent fields, or that they can be retrained into other careers. The skills of an Artist as a judge of quality, aesthetic value, composition, color, and more have varied and broad uses in many fields and in many forms of artwork. Even if a particular niche is replaced by automation, I have no doubt that a talented artist can find work utilizing their well honed abilities.
Or consider how we would organize ourselves collectively in a post-job society (not saying it's arriving tomorrow, but it's certainly something to consider in the future). Whether we want our grandchildren to live it inHunger Games (with the AI company stockowners in the Capitol), Ancient Rome (being a client wasn't that bad, all things considered), The Expanse Earth or Star Trek minus Starfleet will be a political choice we'll have to make.
Misinformation, deep fakes, and illicit uses of imagery is a concern, and AI does accelerate the problem. But it is a problem that has been plaguing the internet for many years.
"News of my death have been greatly exaggerated" predates the Internet... While you're right of course, I hope that the widespread knowledge that "what you see can be false" will help people to redevelop a sense of... skepticality (I don't know the right word) when it comes to information. I've seen flat-earther website and they go to great length to convince you that the Earth is flat; they didn't need AI to do that, but I can see how people are falling for this if they usually trust things that "look serious and well-argumented", at least on the surface. If the future generation are raised on the idea that everything is a deepfake unless it comes from a trusted source and enhance their critical thinking, we may get a good outcome out of it.
