Video explaining their process. It's impressive that they figured out a workflow to achieve consistent and cohesive stylization without it becoming a trippy mess like usual AI generations. Side...
Video explaining their process. It's impressive that they figured out a workflow to achieve consistent and cohesive stylization without it becoming a trippy mess like usual AI generations.
Side tangent: I don't like the overuse of the word "democratization". Sure the techniques are revolutionizing both in terms of productivity and artistic expression. For the average hobbyists though, I'm not sure the transition from having a drawing tablet and an Adobe subscription (if you don't already pirate it) to needing a studio and a high-end computer to train your own AI model is exactly easy on the capital.
While I also don't think there's that much point to talking about "democratization", you can just say "things got easier", but that's not a faithful representation of their argument, which is...
While I also don't think there's that much point to talking about "democratization", you can just say "things got easier", but that's not a faithful representation of their argument, which is mainly about a reduction in labor costs. When you're an amateur or an indie artist, your supply of labor is basically capped - you (and how much caffeine you can ingest). Drawing tablets, and adobe subscriptions, and GPUs are things that can be worked with. But you can't clone yourself, and eventually you can't consume any more caffeine.
A single employee in the US's annual wage + healthcare benefits would buy you an extremely powerful computer, or a very large amount of AWS credits, and cover an adobe subscription with room to spare for a nicer coffee grinder.
As an aside, for what they're doing specifically, it would be an alternative to rotoscoping, which is a valid form of animation, and also extremely expensive form of animation, so the point is valid, but it's not really the same thing as a fully hand drawn animation. Without AI, to do rotoscoping you'd still need a studio and a camera.
There's a big jump from "you and your friends", which are "free" insomuch as you only care about additional capital as opposed to opportunity cost, and any number of employed workers, money wise.
I know of this video soley because it caused a bunch of uproar - see this high profile response video for one example, or this slightly less high profile response to that response (though, fair...
It's interesting how defensive people seem to get when they see AI art. Personally, I see a future where it's possible for an individual artist to make a fully animated TV show. These videos...
It's interesting how defensive people seem to get when they see AI art. Personally, I see a future where it's possible for an individual artist to make a fully animated TV show. These videos depict it as 'undercutting' existing animators, but I see it as removing bottlenecks and allowing for more creative freedom. They also both claim that AI art has no 'soul', or is just a 'collage', but at the same time are worried about it's success, two ideas which combine to result in insults towards the average person's taste.
There will always be a market for high-quality human-animated TV, but if that market isn't large enough to be sustained with living wages, maybe it shouldn't exist. You aren't entitled to every type of luxury entertainment. While it has a negative impact creatively, and maybe culturally, it's not morally bad.
In general, it feels like these criticisms are tied up in a subconscious focus on capitalism and scarcity. Does it negatively affect artists to make these models under capitalism? Probably, but neither of these videos make it clear that's where their criticism is coming from, instead shifting it into a moral issue.
Yeah, I've seen a lot of aggressive criticisms on Twitter that AI tools will allow big studios to make movies without hiring as many animators, but few people seem to consider the converse that...
These videos depict it as 'undercutting' existing animators, but I see it as removing bottlenecks and allowing for more creative freedom.
Yeah, I've seen a lot of aggressive criticisms on Twitter that AI tools will allow big studios to make movies without hiring as many animators, but few people seem to consider the converse that smaller groups with more interesting/experimental ideas will be able to make projects that compete with the more risk-averse bigger studios. That seems much more valuable to me than preserving the practices and labor make-up of large studios.
This guy makes a few relevant points, but his considerations about the evolution of the art form are extremely luddite and misinformed. AI won't remove the human element from art, it will merely...
I know of this video soley because it caused a bunch of uproar - see this high profile response video for one example
This guy makes a few relevant points, but his considerations about the evolution of the art form are extremely luddite and misinformed. AI won't remove the human element from art, it will merely transform it. It's another brush.
Generally speaking, any anti AI art argument focusing on the purity and essential nature of the art forms is bound to be fraught, emotional, weak, and, to me, unpersuasive. The strong arguments are the ones about the ethical and economic aspects. The worthy discussion is not "is this art?" (because it obviously is), nor "is this good?" (because even if it isn't, it obviously will be), but rather "is this fair?".
Finally, this is rotoscoping. So, if on one hand, it is reasonable to worry about animators' jobs, on the other many actors and actresses will find employment with this technique.
The thing about the Luddites is that they weren't wrong. Don't take this as a personal attack, but I feel that you've missed the point in their criticisms. You dismiss these arguments as being...
Don't take this as a personal attack, but I feel that you've missed the point in their criticisms. You dismiss these arguments as being emotional, but when it comes to art, emotion is the point. And it's fine if you aren't persuaded by it because art is also famously subjective. At the same time, you really should treat those criticisms with at least a bit of respect because those subjective things obviously matter very much to those people.
(For the record I did get a chance to watch that second video and I do agree that the speaker in it does have some inaccurate statements, though I don't feel that those invalidate his greater points.)
I also don't think that you are right about AI animation ever becoming as good as humans are at traditional 2D animation. Some things will improve, for certain; the irregular lines, choppy irregular shapes flickering in and out of existence, the weird glitchy spots, and even the terrible hands might all go away. But when I compare the video they produced to some of the clips from the source material they used to train their AI and the difference is night and day. It's not just that the originals are smoother, more regular, more "on model" or anything like that, it's that there's a lot more effort put into the overall composition of the work; the framing, the coloring, and even the aesthetics of motion are all consciously planned in meticulous detail before anyone even begins work on the frameworks of the frames that will go into the final picture. To be clear, I think that AI may be able to do this kind of work in the distant future, but right now the tech isn't nearly that advanced, and I doubt that we'll see that kind of thing happen even in our lifetime.
Yes, I know we are seeing a huge amount of growth in AI in the past few years, but we really don't know how sustainable that growth is. There was a time when people thought unironically that Moore's Law was going to continue indefinately. Assuming that AI will be the same way is another example of wishful pop-futurism. In reality it's far too early to tell what will happen.
Anyone who understands the realities of producing animation knows that AI tools are inevitable. There are already productions that have been released that use AI-assisted tweening. What people are upset about is the idea that the human aspects of creating their art disappearing. And that is something that I don't think will ever truly happen. There may be an AI that can make a decent approximation of an artistic work sometime in the unforseeable future, but people intrinsically value the efforts of other human beings.
You did not offend me in the slightest, but I think your interpretation of my comment is incorrect. I understand that emotion is essential to the aesthetic experience which most people associate...
You did not offend me in the slightest, but I think your interpretation of my comment is incorrect.
I understand that emotion is essential to the aesthetic experience which most people associate with art. When I used the word "emotional" I didn't mean to deny that, but rather to say that those arguments themselves are emotional in a way that does not contribute to the persuasiveness of their criticism in this particular case. The emotional features of arguments about art are not the same as the emotional results that art produces through its fruition. Those are very different things that I think you mistakenly equated.
Ultimately, what we need is human connection and the feeling that whatever we are consuming expresses some aspect of existence to which we connect on a fundamental level. My point is that I don't see that changing anytime soon, and I must be very clear that I'm talking specifically about the case that is being discussed in this post. The video was created by humans, including two real actors, using AI tools. They wrote the story, staged it with actors, curated the dataset, manipulated the AI to give the results they wanted, and edited it all together relying on their own predictions and expectations of its effects on human beings. AI is a brush.
It doesn't matter to me that they didn't draw the actual frames, they are the ones in charge of producing something that is meaningful to humans.
That is both awful and impressive, and, to me, potentially more pleasing than human-made 3D-CGI anime. For now, I see it being useful as one step of the process, with animators manually fixing the...
That is both awful and impressive, and, to me, potentially more pleasing than human-made 3D-CGI anime.
For now, I see it being useful as one step of the process, with animators manually fixing the problems and giving the final touches. Eventually, the word "animator" might be how we call a specific animation AI, in the same way that "computers" used to be real people.
Video explaining their process. It's impressive that they figured out a workflow to achieve consistent and cohesive stylization without it becoming a trippy mess like usual AI generations.
Side tangent: I don't like the overuse of the word "democratization". Sure the techniques are revolutionizing both in terms of productivity and artistic expression. For the average hobbyists though, I'm not sure the transition from having a drawing tablet and an Adobe subscription (if you don't already pirate it) to needing a studio and a high-end computer to train your own AI model is exactly easy on the capital.
While I also don't think there's that much point to talking about "democratization", you can just say "things got easier", but that's not a faithful representation of their argument, which is mainly about a reduction in labor costs. When you're an amateur or an indie artist, your supply of labor is basically capped - you (and how much caffeine you can ingest). Drawing tablets, and adobe subscriptions, and GPUs are things that can be worked with. But you can't clone yourself, and eventually you can't consume any more caffeine.
A single employee in the US's annual wage + healthcare benefits would buy you an extremely powerful computer, or a very large amount of AWS credits, and cover an adobe subscription with room to spare for a nicer coffee grinder.
As an aside, for what they're doing specifically, it would be an alternative to rotoscoping, which is a valid form of animation, and also extremely expensive form of animation, so the point is valid, but it's not really the same thing as a fully hand drawn animation. Without AI, to do rotoscoping you'd still need a studio and a camera.
There's a big jump from "you and your friends", which are "free" insomuch as you only care about additional capital as opposed to opportunity cost, and any number of employed workers, money wise.
I know of this video soley because it caused a bunch of uproar - see this high profile response video for one example, or this slightly less high profile response to that response (though, fair warning, I haven't had the chance to watch this one yet, myself).
It's interesting how defensive people seem to get when they see AI art. Personally, I see a future where it's possible for an individual artist to make a fully animated TV show. These videos depict it as 'undercutting' existing animators, but I see it as removing bottlenecks and allowing for more creative freedom. They also both claim that AI art has no 'soul', or is just a 'collage', but at the same time are worried about it's success, two ideas which combine to result in insults towards the average person's taste.
There will always be a market for high-quality human-animated TV, but if that market isn't large enough to be sustained with living wages, maybe it shouldn't exist. You aren't entitled to every type of luxury entertainment. While it has a negative impact creatively, and maybe culturally, it's not morally bad.
In general, it feels like these criticisms are tied up in a subconscious focus on capitalism and scarcity. Does it negatively affect artists to make these models under capitalism? Probably, but neither of these videos make it clear that's where their criticism is coming from, instead shifting it into a moral issue.
Yeah, I've seen a lot of aggressive criticisms on Twitter that AI tools will allow big studios to make movies without hiring as many animators, but few people seem to consider the converse that smaller groups with more interesting/experimental ideas will be able to make projects that compete with the more risk-averse bigger studios. That seems much more valuable to me than preserving the practices and labor make-up of large studios.
This guy makes a few relevant points, but his considerations about the evolution of the art form are extremely luddite and misinformed. AI won't remove the human element from art, it will merely transform it. It's another brush.
Generally speaking, any anti AI art argument focusing on the purity and essential nature of the art forms is bound to be fraught, emotional, weak, and, to me, unpersuasive. The strong arguments are the ones about the ethical and economic aspects. The worthy discussion is not "is this art?" (because it obviously is), nor "is this good?" (because even if it isn't, it obviously will be), but rather "is this fair?".
Finally, this is rotoscoping. So, if on one hand, it is reasonable to worry about animators' jobs, on the other many actors and actresses will find employment with this technique.
The thing about the Luddites is that they weren't wrong.
Don't take this as a personal attack, but I feel that you've missed the point in their criticisms. You dismiss these arguments as being emotional, but when it comes to art, emotion is the point. And it's fine if you aren't persuaded by it because art is also famously subjective. At the same time, you really should treat those criticisms with at least a bit of respect because those subjective things obviously matter very much to those people.
(For the record I did get a chance to watch that second video and I do agree that the speaker in it does have some inaccurate statements, though I don't feel that those invalidate his greater points.)
I also don't think that you are right about AI animation ever becoming as good as humans are at traditional 2D animation. Some things will improve, for certain; the irregular lines, choppy irregular shapes flickering in and out of existence, the weird glitchy spots, and even the terrible hands might all go away. But when I compare the video they produced to some of the clips from the source material they used to train their AI and the difference is night and day. It's not just that the originals are smoother, more regular, more "on model" or anything like that, it's that there's a lot more effort put into the overall composition of the work; the framing, the coloring, and even the aesthetics of motion are all consciously planned in meticulous detail before anyone even begins work on the frameworks of the frames that will go into the final picture. To be clear, I think that AI may be able to do this kind of work in the distant future, but right now the tech isn't nearly that advanced, and I doubt that we'll see that kind of thing happen even in our lifetime.
Yes, I know we are seeing a huge amount of growth in AI in the past few years, but we really don't know how sustainable that growth is. There was a time when people thought unironically that Moore's Law was going to continue indefinately. Assuming that AI will be the same way is another example of wishful pop-futurism. In reality it's far too early to tell what will happen.
Anyone who understands the realities of producing animation knows that AI tools are inevitable. There are already productions that have been released that use AI-assisted tweening. What people are upset about is the idea that the human aspects of creating their art disappearing. And that is something that I don't think will ever truly happen. There may be an AI that can make a decent approximation of an artistic work sometime in the unforseeable future, but people intrinsically value the efforts of other human beings.
You did not offend me in the slightest, but I think your interpretation of my comment is incorrect.
I understand that emotion is essential to the aesthetic experience which most people associate with art. When I used the word "emotional" I didn't mean to deny that, but rather to say that those arguments themselves are emotional in a way that does not contribute to the persuasiveness of their criticism in this particular case. The emotional features of arguments about art are not the same as the emotional results that art produces through its fruition. Those are very different things that I think you mistakenly equated.
Ultimately, what we need is human connection and the feeling that whatever we are consuming expresses some aspect of existence to which we connect on a fundamental level. My point is that I don't see that changing anytime soon, and I must be very clear that I'm talking specifically about the case that is being discussed in this post. The video was created by humans, including two real actors, using AI tools. They wrote the story, staged it with actors, curated the dataset, manipulated the AI to give the results they wanted, and edited it all together relying on their own predictions and expectations of its effects on human beings. AI is a brush.
It doesn't matter to me that they didn't draw the actual frames, they are the ones in charge of producing something that is meaningful to humans.
Ah, I see what you mean now.
That is both awful and impressive, and, to me, potentially more pleasing than human-made 3D-CGI anime.
For now, I see it being useful as one step of the process, with animators manually fixing the problems and giving the final touches. Eventually, the word "animator" might be how we call a specific animation AI, in the same way that "computers" used to be real people.
And the voice acting is shit.
To think