PCG Article Class Action Lawsuit filed by artists against AI Image Generators

Page 2 - Love gaming? Join the PC Gamer community to share that passion with gamers all around the world!
I'm pretty sure a lot of people who generate images with AI tools have no idea how those tools work or that those tools might have been trained on copyrighted material. I'm pretty sure a lot of the users therefore aren't making an informed decision about using someone else's work.
I'm not so sure and at the moment I'm trying to resolves multiple ways of undersatnding this. I think anyone who is using it is aware of what it's doing, ie, scanning the Web for any images that relate to the inputed text. And each time they add more text, or the name of an artist they can see how it's using what it finds and the image is changing, it's happening almost in real time in front of them. I can see how impressive that is.

Do you remember the whole sampling debate, at first musicians just took samples of other people's music and from those constructed something new. But then musicians made it so you could only take their creations and use it if you had permission, if you didn't then all royalties made went to the original artist. The Verve's Bitter Sweet Symphony is the well known one, where 100% of royalties went to the Rolling Stones, although apparently the Stones have now reversed that. But you often hear of lawsuits brought by musicians that think there music has efectively been stolen without their permission.

At the moment with say Midjourney, it is viewed as the creator and if you do sell work then Leap Motion(I believe the company is called) take 20%.

I think the main concern of artists, is they train for decades, they have the talent and say the vision and someone else can just 'steal' that. It's like the robots taking manual jobs, AI is taking intellect/talent jobs.

Art generating AI will change the way the creative industries work, from wall prints to gaming.

But the AI works that are made from scanning other people's works doesn't at present seem to have that human element to it, that depth, and of course that makes sense, that's the level it's operating at, whether that will change I don't know.

Artists must be concerned that if their works are being used like this it may make many redundant from a wide range of creative industries. And they have every right to protect their unique creations and livelihoods.

One of my concerns is that art as culture is continually developing, evolving and progressing and incorporates different aspects of what it is to be human from a philosophical and psychological viewpoints. And human audiences respond to that, it can often make sense of what it is to be a human at any given time.

This new AI art generation may be just a phase, a novelty, but it could also change art into a continually self referencing, self replicating sort of simulation loop. Who Knows!
 
I'm pretty sure a lot of people who generate images with AI tools have no idea how those tools work or that those tools might have been trained on copyrighted material. I'm pretty sure a lot of the users therefore aren't making an informed decision about using someone else's work.
Yeah, I guess you're right. I was thinking of the ones where you have to give the AI a few pictures to work off of. I forgot they're not all like that.
 

Zloth

Community Contributor
To really make a mess of things, what happens when we start mixing things up? AI, sing the Canadian national anthem with Brittney Spears' voice in the style of Elvis Presley.

Want to copyright voice? Fine, pitch Spears' voice down an octave and add a Scottish accent. Style? Do a painting of a space station with the brush strokes of this dude and the color choices of that dude.

(Hmmm... AI, do this Wheezer video but have Buddy Holly sing it and replace the images with Buddy Holly and the Crickets.)
 
Like your thinking, it makes me think of post modernism roulette(especially; randomness[Zen influence], playfullness, cross referencing, etc). And who doesn't like Wheezer!

I enjoy doing that, take a sample, slow it down, reverse it, pitch shift, etc basically just play with it until the creator wouldn't stand a chance without an AI.

johnny cash elvis <an impersonation of a rock and roll singer impersonating Elvis
 
To really make a mess of things, what happens when we start mixing things up? AI, sing the Canadian national anthem with Brittney Spears' voice in the style of Elvis Presley.

Want to copyright voice? Fine, pitch Spears' voice down an octave and add a Scottish accent. Style? Do a painting of a space station with the brush strokes of this dude and the color choices of that dude.

(Hmmm... AI, do this Wheezer video but have Buddy Holly sing it and replace the images with Buddy Holly and the Crickets.)
Lol. When you spelled it "Wheezer," I thought you were talking about this video...

View: https://www.youtube.com/watch?v=o4F3J5zFM8M
 
Art generating AI will change the way the creative industries work, from wall prints to gaming.
I think this is a definite possibility, at least years down the road, because I don't think this is a tech that is going to just fade away and may eventually branch out beyond the arts into other parts of our lives. Looking at these AI programs in a broader sense, I think there's an issue with individuals not being able to "opt-out" of having their work, whatever it is, being vacuumed into these huge databases and used by anyone. Denying individuals a choice is never a good idea in my opinion.

Artists must be concerned that if their works are being used like this it may make many redundant from a wide range of creative industries. And they have every right to protect their unique creations and livelihoods.
This is one of my major sticking points with this new AI technology. While I'm interested in how this tech evolves, especially when it comes to gaming AI, I don't want it to be at the expense of individual artists.

For example, let's say @ipman has his work included (without his permission) into an AI database. Another individual uses assets from his work and creates other images, then makes thousands of dollars by selling those images to an advertising agency, or from prints shown in a gallery. Doesn't IP have a say in allowing that, getting a credit, or receiving some of the profit?

I think these branches over into the example that @Zloth made about music/musicians, where someone's voice or music is being altered and produced as something else.
 
I think this is a definite possibility, at least years down the road, because I don't think this is a tech that is going to just fade away and may eventually branch out beyond the arts into other parts of our lives. Looking at these AI programs in a broader sense, I think there's an issue with individuals not being able to "opt-out" of having their work, whatever it is, being vacuumed into these huge databases and used by anyone. Denying individuals a choice is never a good idea in my opinion.


This is one of my major sticking points with this new AI technology. While I'm interested in how this tech evolves, especially when it comes to gaming AI, I don't want it to be at the expense of individual artists.

For example, let's say @ipman has his work included (without his permission) into an AI database. Another individual uses assets from his work and creates other images, then makes thousands of dollars by selling those images to an advertising agency, or from prints shown in a gallery. Doesn't IP have a say in allowing that, getting a credit, or receiving some of the profit?

I think these branches over into the example that @Zloth made about music/musicians, where someone's voice or music is being altered and produced as something else.
I don't think it will take long before companies and gaming companies start using this tech. Not that long ago artists were employed to completely draw all the characters, costumes and machines for a game like HZD.

I remember looking at the original concept drawings, by Jan Bart van Beek, and then an art team worked up those drawings, see>

https://magazine.artstation.com/2017/05/guerrilla-games-horizon-dawn-art-blast/

But it will be again down to cutting costs if dev companies don't need to employ those people.

Also with such interest and the possibilities I think these AI generators will develop quickly.

So from the artists point of view it could be disasterous.

Sampling is slightly different but also similar. Sampling music seems to back to the 1940's at least but did really become mainstream in the 1970's with Hip Hop, but it wasn't until a court case in 1991 that samplers had to get permission first. But as I mentioned elsewhere sampling The Winstons Amen break gave rise to a new genre.

The problem is so much art work is out there on the Web already. And although it took law changes and copyright protection a long time to catch up with sampling, this AI generation tech is moving so quick I can see why artists are trying to protect their works as quickly as possible.

I think as you say artists should have the right to 'opt out' but it may be to late. But I think if some one is going to use an artists creations they should have to get permission, and then pay some sort of royalty.

It's interesting though that the AI generator is seen as the 'creator', and the companies that own them can demand royalties. That in many ways makes it easier for the artists, they may be able to claim the company asks permission and has to make a deal to share some of that 20% the companies are taking.
 
I think as you say artists should have the right to 'opt out' but it may be to late. But I think if some one is going to use an artists creations they should have to get permission, and then pay some sort of royalty.

It's interesting though that the AI generator is seen as the 'creator', and the companies that own them can demand royalties. That in many ways makes it easier for the artists, they may be able to claim the company asks permission and has to make a deal to share some of that 20% the companies are taking.

In the EU it's already possible to opt out. Also, it seems that in the US, AI-generated art cannot be copyrighted in most circumstances.
 
Apr 6, 2023
1
1
15
Visit site
I don't think the lawsuit is going anywhere. In a situation where every day there are dozens of tools to create content, photos, and drawings ... using AI technology are created, I think artists and photographers should learn to use them as a tool. the ancillary tool rather than going about pointless lawsuits. With their ability, I think true works of art will be born a thousand times faster than before.
 
  • Like
Reactions: Brian Boru

TRENDING THREADS