The AI Red Scare is only harming artists and needs to stop.

[bedir than]

I'm not sure how they did it but the key point is that as far as I know they obtained all the data legally.

OpenAI isn't willing to make the claim about their technology that you are.
They have insisted that they will never reveal their sources nor how they sourced it. They have also stated that if Copyright law in the US applies to them they wouldn't be able to exist.

I trust Altman and OpenAI to understand themselves better than someone on a TTRPG board.

 

[LordEntrails]

(nothing here)

 

[Celebrim]

OpenAI isn't willing to make the claim about their technology that you are.

Oh?

They have insisted that they will never reveal their sources nor how they sourced it.

That's not a contradictory statement. I'm not going to reveal proprietary information about my company either. My unwillingness to do so is not a confession of wrongdoing.

They have also stated that if Copyright law in the US applies to them they wouldn't be able to exist.

And also, that statement doesn't mean what you seem to think it does. I agree that if copyright law in the USA applies to large language model creation then they wouldn't be able to exist, but also note that that is a conditional statement. The second part depends on truth of the first part. The point is, copyright law doesn't apply to what they did, and I suspect any cases brought against them will be resolved in their favor because the judges will agree. And so far, that's what all the judges have found. For example, OpenAI successfully defended itself from writers like Sarah Silverman who argued that she should have been financially compensated for OpenAI processing her memoir as input to their model.

You see the truth is that OpenAI is not admitting or claiming it was engaged in any wrongdoing, quite to the contrary of the innuendo of your statements.

 

[Golden Bee]

inevitable AI revolution

I love that word, "Inevitable." It doesn't solve the problem, but it sure makes it sound like it's someone else's problem. There's nothing we can do about it, it's inevitable. It would have happened anyway, it's inevitable. It was only a matter of time. Eventually it would have sprung up from the ground fully-formed, all by itself. It's inevitable.

But it isn't. It was created by specific people, doing specific things, for specific reasons.

It is true that AI technology is here to stay, and it isn't going away...but that's not what "inevitable" means. You're thinking of the word "irrevocable." But that lacks impact. It doesn't set the right tone.

"Inevitable" sounds cool, like an edgy comic book supervillain making an astute observation about the state of society, people get philosophical and say "hmm yes, I see." "Irrevocable" sounds bureaucratic, like a lawmaker trying to sneak something onto a ballot, people poke their heads up and say "oh? is that so?"

 

[Necropolitan]

You see the truth is that OpenAI is not admitting or claiming it was engaged in any wrongdoing, quite to the contrary of the innuendo of your statements.

They claimed their actions don't violate copyright law AND said they couldn't create AI without using copyrighted works.

The first claim is very much in dispute.

OpenAI is in fact very obviously violating copyright, which is why they get hit with lawsuits the moment the copyright holder gets proof they're using their intellectual property.

As The New York Times pointed out OpenAI outright copied their articles for AI training. That's blatant infringement. Even worse OpenAI used the articles to create a competitor to The New York Times.

And the fact AI creators can't create AI without using copyrighted works means that if it's decided that AI creators ARE violating copyright AI research is going to hit a wall.

 

[Lanefan]

An AI isn't 'someone,' it's a program that copies the entire image and then regurgitates parts of it.

I could memorize a picture in a gallery, go home, and regurgitate part of it on my own canvas*.

Same process, same result, only more work for me and it takes longer.

* - were I a better artist; for me it'd be music, where I know someone else's song well enough to incorporate maybe-or-maybe-not recognizable bits of it into my own song along with bits of other songs and my own additions.

 

[Necropolitan]

I could memorize a picture in a gallery, go home, and regurgitate part of it on my own canvas*.

Same process, same result, only more work for me and it takes longer.

* - were I a better artist; for me it'd be music, where I know someone else's song well enough to incorporate maybe-or-maybe-not recognizable bits of it into my own song along with bits of other songs and my own additions.

You're not literally copying the entire thing and then outputting part of it.

Do you not know the difference between inspiration and plagiarism?

 

[Lanefan]

No, in my case, quite the contrary. Digital piracy is obviously theft. The people going "If purchasing isn't owning then piracy isn't theft" are moral cretins that are trying to justify theft. Even if I can see why you might worry about the digitization of information and the lack of physical copies, that still doesn't justify theft. But no piracy occurred in this case because no permanent copy of the information was made. @CapnZapp 's analogy doesn't work because it doesn't represent what actually happened. (And I again protest that analogies don't bring clarity, but only confusion.) What really happened was more like a student in a trade school opening the hood of a car to see how cars worked, then giving it back intact to the owner. But again, I don't think analogies are actually helpful.

In reality, the only copy that was made was the temporary copy that is made when anything on the internet is viewed, which is an essential aspect of the technology without which the whole internet must be taken down. Do you understand how the internet works?

I understand how the "Save image..." command works, which makes that "temporary" copy permanent at least within my computer and truly permanent if I decide to print it out.

Everyone is influenced by everything that they have ever read, but that doesn't mean you have to cite everything you ever read. John Williams is influenced by everything he ever heard, but isn't required to list all of his inspirations in his liner notes. Transformative work is transformative work.

I think the point being made by some is that AI art is also just transformative work.

Question for the AI detractors: if AI only used public-domain material to train on, would that make it acceptable?

 

[Necropolitan]

Question for the AI detractors: if AI only used public-domain material to train on, would that make it acceptable?

As multiple people have pointed out the copyright infringement is a major issue.

And it being trained on public-domain material wouldn't change the fact that AI is plagiarism.

 

[Lanefan]

You're not literally copying the entire thing and then outputting part of it.

Maybe I'm in fact doing exactly that, only the "copy" I've made is in my own memory rather than on a piece of paper or in some form of digital storage.

Do you not know the difference between inspiration and plagiarism?

At what point does one become the other?