The issue with AI art (and especially with AI code as well) is that it’s just spicy plagiarism. The AI just looks through millions of works of art (or trillions of lines of code) and spits out something based on that. There’ve been cases of GitHub Copilot spitting out someone else’s code verbatim. Fuck RIAA, fuck the copyright, but it’s not a clear cut case.
To be fair, this applies to humans the same way. Art and inventions are derivative of prior work. Artists study existing techniques and styles when they develop their own. There are tons of examples of musicians reusing existing riffs and melodies, and so on.
Same, I kind of see AI art generation to be akin to photography. A photographer doesn’t create the scene, they merely have an eye for capturing it. It could be argued that someone generating images using a machine learning algorithm and getting the algorithm to produce a picture they find interesting is essentially doing the same task.
I do see this becoming a tool for artists as well. For example, it’s possible to give stable diffusion a sketch and have it fill in details in the sketch using prompts. This saves a lot of manual work for the artist. The artist can then build on this. This can also be used as a prompt for ideas where you’d generate a bunch of pictures and then use one as inspiration for a picture of your own.
As I see it, the real question is what role we think mechanical skill plays in creation of art. For example, modern tools like Krita make it much easier for artists to produce intricate effects than a traditional medium. It takes far more skill to paint with oil on canvas than to produce digital art. Does it mean that makes a picture painted using digital tools somehow lesser or is it the vision of the artist that ultimately matters. And if we agree that it comes down to the vision and the ideas or feelings the artists aims to convey then the medium should not matter.
Furthermore, the observer imparts their own meaning on the art. When I look at a picture, it’s either meaningful to me or not. It hardly matters whether it was produced by an algorithm or by a person. In most cases I have no idea what the artist was thinking when they painted a picture, so any meaning I ascribe to it comes solely from me.
@yogthos nope, the difference is technique is just a means of saying some thoughts. So people study others technique to be more shiny in the world of thoughts. And AI doesn’t produce any new thoughts. That’s the difference.
100 monkeys in a room could make music as good as what our company publishes. Regulation is needed for the congregating of monkeys.
The issue with AI art (and especially with AI code as well) is that it’s just spicy plagiarism. The AI just looks through millions of works of art (or trillions of lines of code) and spits out something based on that. There’ve been cases of GitHub Copilot spitting out someone else’s code verbatim. Fuck RIAA, fuck the copyright, but it’s not a clear cut case.
To be fair, this applies to humans the same way. Art and inventions are derivative of prior work. Artists study existing techniques and styles when they develop their own. There are tons of examples of musicians reusing existing riffs and melodies, and so on.
Some of this backlash might definitely be just artists/developers afraid of competition. But it’d be interesting to see how the space develops.
Same, I kind of see AI art generation to be akin to photography. A photographer doesn’t create the scene, they merely have an eye for capturing it. It could be argued that someone generating images using a machine learning algorithm and getting the algorithm to produce a picture they find interesting is essentially doing the same task.
I do see this becoming a tool for artists as well. For example, it’s possible to give stable diffusion a sketch and have it fill in details in the sketch using prompts. This saves a lot of manual work for the artist. The artist can then build on this. This can also be used as a prompt for ideas where you’d generate a bunch of pictures and then use one as inspiration for a picture of your own.
As I see it, the real question is what role we think mechanical skill plays in creation of art. For example, modern tools like Krita make it much easier for artists to produce intricate effects than a traditional medium. It takes far more skill to paint with oil on canvas than to produce digital art. Does it mean that makes a picture painted using digital tools somehow lesser or is it the vision of the artist that ultimately matters. And if we agree that it comes down to the vision and the ideas or feelings the artists aims to convey then the medium should not matter.
Furthermore, the observer imparts their own meaning on the art. When I look at a picture, it’s either meaningful to me or not. It hardly matters whether it was produced by an algorithm or by a person. In most cases I have no idea what the artist was thinking when they painted a picture, so any meaning I ascribe to it comes solely from me.
@yogthos nope, the difference is technique is just a means of saying some thoughts. So people study others technique to be more shiny in the world of thoughts. And AI doesn’t produce any new thoughts. That’s the difference.
@Cysioland
See my other reply where I address this.