Not everything is bleak. The ending sequence for Frieren season 2 was entirely hand drawn with colored pencils[0] and the new Ghost in the Shell (coming out next month) is also hand drawn[1].
Neat, I didn't know for Ghost in the Shell, I'll be watching it in few weeks then! Thanks for sharing.
PS: so weird to watch the trailer while working in XR. I never imagined as a kid I'd be programming in a headset but now it's banal. I even buy 2nd hand HMDs for 100 bucks. Weird times.
Noob Question: There is a famous parlor trick with generative networks(I think it was GANs but it might be some kind of diffusion based network.), you start with a canvas and draw a stick figure of what you want and the generative network draws the rest of it.
Do AI platform companies actually pre-train networks to do the same for hand drawn artists?
Related question: If they do train them to do that, are there any that train people for the "reverse": learn how to draw with paper and pencil by showing techniques only i.e only the "what" but not the "how" ?
Even if we're still far from the golden age, Dandadan (Science SARU; though I've heard the sheer effort bankrupted them) and the recent Witch Hat Atelier (Bug Films) are quite nice to behold.
Some modern romcoms like Kaguya-sama (A-1 Pictures) and The Dangers in My Heart (Shin-Ei) were nicely done too.
Things are obviously not going in the right direction, but the current accelerated fall in quality the consumer sees is more influenced by general unwillingness to spend and/or take risks from bean counting production committees.
After reading this article, I got curious whether there was a similar article from Japan, and there actually is. The average monthly income in the animation industry is 200,000 yen (about $1,300 USD), but the median working hours are 2,745 hours per year. That comes out to 225 hours per month, or 52 hours per week [1][2]. Considering that animators' work is essentially drawing labor, that's an insane amount of work. But even as total production costs and promotional effects grow, none of it reaches the workers on the ground. It seems like in modern industry, the value of promotion and fame outweighs what the laborers actually produce.
Actually, when you think about it, this problem is happening across all sectors of society. Ultimately, it's a system where platforms intermediate and monopolize value.
Platforms concentrate their investment in IP and star creators, and the commercial success of these creators in turn increases the platform's value, creating a virtuous cycle. However, this success ultimately ends up concentrated among a small upper tier, while the vast majority are excluded.
The article essentially says the same thing.
It seems like we're in the age of platform capitalism. Come to think of it, the programming world feels similar too
There's a japanese docuseries called 'Manben' (available on youtube) that interviews a bunch of famous manga artists. Without fail they all talk about how much work, hours, suffering goes into the first years of making manga.
It's very interesting to watch and I highly recommend it. But it's also a GREAT advertisement for avoiding the industry.
So basically they get paid peanuts and are overworked and so no one wants to do it.
Also seems like something AI could really cut into. You could have a master animator doing much of important thoughtful work and AI filling in the obvious as well as doing tweening (sound similar to programming)
Really, who needs a studio if you have the creative talent and ability to leverage AI for the grunt work. Or have a couple grunt work humans paid these rates to manage the tedious work of leveraging the AI to make it look seamless.
AI image generation is just not there yet. Say what you want about Luddites and whatnot, but the quality is just not good - the amount of effort given compensation that (especially Japanese) artists put into anime styled imagery makes too little sense that AI can't compete, even in generation time if time for retakes are accounted for.
You're not wrong. But there is a common perception that we value things made by humans more. The problem is that grunt work actually serves as a pipeline for industrial training. Even with AI, the distribution of value doesn't get resolved automatically.
Of course, I think it would be great if grunt work disappeared, but I believe skilled workers ultimately need grunt work. It's like saying that since AI automates everything, programmers don't need to know how to write methods. The core issue here is that grunt work, which AI excels at, plays an educational role in our society.
Of course, I admit my thinking is quite old-fashioned. This educational model could change. But I'm not sure whether that would be good in the long run. It could be beneficial in the long term. Humans evolve, after all.
> Of course, I think it would be great if grunt work disappeared, but I believe skilled workers ultimately need grunt work. It's like saying that since AI automates everything, programmers don't need to know how to write methods. The core issue here is that grunt work, which AI excels at, plays an educational role in our society.
It's not just educational. The more thinking you offload to AI, the more your own skills degrade [1] - and it makes sense, intuitively. If you repeat tasks, you gain experience and get good at it... but if you cease that repetition, eventually your skills break down.
I've thought a lot about how to characterize the difference between 3D and hand-drawn stuff. I think the core of it is:
- With hand drawn animation, you draw what you have in your head, first roughly, then you refine it
- With 3d animation, you first need to model everything, then rig it, then work with the bone system to get the motions you want or else mocap, and then set up rigs and stuff so that the mocap actors can do crazy movements, etc etc. Then maybe undo some of the scaffolding the 3d software does: disconnect bones, fake perspective. You have to fiddle with lighting, textures, etc. Or you don't, and just go with whatever's easy to do in the 3d software.
Which means that spontaneity and emotion, like I think this guy's arms should be all wiggly here, are lost. Yeah, you can hand animate then 3d animate on top of the hand animation, but in an industry that's using 3d to cut costs and not because it looks better, that's not going to happen (in any way that keeps the spontaneity).
3d is awesome in that once you've done a huge amount of up-front prep, the rest is easy to iterate on and tweak, but that's a large tradeoff.
I thought that this is one area where I think AI could be a force of good. Keep the animators doing the rough sketches, and AI comes up with the lines, handles the filling, and maybe adds colors with a guide. I haven't seen this yet.
---
I'm not sure I agree on the mentorship parts. IIUC all the major studios and famous animators weren't taught by someone. All the studios have unique flairs that they came up with just by playing around and copying Disney. And they got there without drawing hundreds of thousands of in-between frames for someone else.
I think that being taught the correct way to animate based on existing productions probably also reduces creativity in the field.
I wonder if the earlier creativity was due to voids though, and now that there's some amount of saturation it's harder to break in, or if somehow the increase in revenue from global interest somehow increases stakes and causes more downward pressure squashing out experimentation.
---
There's a lot I don't get about this article though. It says the demand is way up, but the treatment and pay for animators is terrible... why? I didn't see it addressed. Japan has a long history of "non-monetization" though, like refusing to sell digital music overseas, or regional restrictions on streaming content.
---
Lastly, I think there's still a lot of indie animation that gets glossed over. There are lots of independent animators making animations for music videos, for instance, or releasing small animations. I don't know if that grows into larger productions, but there's a level of creativity you'd never see if you just watch televised anime.
Interesting, the "training of new animators" mirrors what has/is happening in other industries.
When I started programming decades ago, an experienced programmer would review my work and help me out. That started ending in the very late 80s and 90s. By 2000 or so, you were on your own as a new employee. I even mentioned it to a high level manager a while ago, he said we expected people we hire to know what they are doing.
I have heard similar things have occurred in manufacturing too.
Remember the joke about asking for 5 years of experience with a 5 month old framework?
This is part of the whole move fast and break things mantra. If you have to train people you aren't moving fast enough. And now they can bolt on AI turbochargers.
It'll be very interesting reading future studies on how this has negatively impacted entire generations. Hopefully people realize that you need to pay younger people living wages to learn skills, if they want people to have those skills in the future.
My (probably unpopular here) opinion is just the opposite. We need more of an apprenticeship model where you're not paid a bunch because you're still learning, and you probably bring negative-to-zero value initially. When a fresh-out-of-college junior engineer brings in SV style money, the expectation should be that they already know what they're doing.
In the trades you start off low pay because you're generally more in the way than helpful, but then you gain the experience and knowledge to be valuable.
Even the resident-doctor relationship is like this. Resident are overworked and poorly paid because they are more distracting than the value they add, then there's the big reward at the end.
The grad-student/professor model is kind of like this too except for all the pyramid scheme stuff that happens there.
I think most technical fields need to go to this model where the newbie commits to learning and trying to be valuable instead of rest-and-vest. And then once they're valuable they get paid more in proportion to the value they bring to the field.
In my small company we had to switch about 5 years ago to only hiring folks with lots of experience (10-15 years). We tried hiring younger fresh-out-of-college engineers, but "market rate" was too high and they required too much attention from senior staff and it made us unsustainably unproductive. We wanted to mentor and teach the next generation, but we couldn't afford it.
i love an apprenticeship model, i think most people would learn better and become competent at whatever position much quicker than straight book learning, for practically every job.
however, where you say: "My [...] opinion is just the opposite [...] where you're not paid a bunch ", are you saying the opposite of a living wage? how would you expect someone to, well, live during their apprenticeship? someone starving and worried about getting evicted or similar is not in a great head space to learn effectively.
Let in the chronicles be written, that widespread inability to read and write endeth the chronicles.. class of mistake. Let this be a lesson to all readers of chronicles.. the big failure starts three generations before the white paper appears..
> we expected people we hire to know what they are doing
I feel this is a generational thing. Many baby-boomer parents never took the time to teach their children any skill. They thought they learnt it by osmosis I guess.
Their generation outsourced everything they could.
Anime has never been more popular but honestly its quality has declined a lot. It's all isekai and supernatural-sexy-schoolgirl-whatever (kinda creepy) these days. Or the same old animes that go on forever.
Every year there's less and less animes that are worth the time to watch IMHO.
The same ratio of garbage to quality was available back then, there just wasn’t the same level of exposure to the range of output so the worst ones faded into history.
Funny, that was already true in 2006. It's the thing people keep saying, and yet anime keeps coming out.
Right now they're stuck in the whole "ten shows from the same budget, each run by their own 'committee' and each competing for cash that runs out well before the show's over, and the poor performance very quickly get less per episode". Great for networks, shit for shows. Even worse for animators who need to get paid a living wage.
A cursory glance of anime tv shows released in 2006 and there's a lot of good things I'd watch again: Honey & Clover, Kemonozume, Ergo Proxy, Code Geass, Haruhi Suzumiya, etc.
In 2026 the only thing that has captivated me, so far, was the Chainsaw Man movie and Dorohedoro S2.
Higurashi when they cry does the whole concept of time reset well (2006).
Trigun sequel is coming out this year.
2026 titles have caught on that trend of a really long specific title though.
“The Laid-Off Cheat-Granting Mage Enjoys a Second Lease on Life”
I think anime is a big enough category that every year there’s always at least a few good ones but it’s hard to get good signal to noise. But the whole tiny girl thing is 100% a trend that’s stuck around.
Maybe there are only two things you would watch but there are tons of others that a lot of people are watching. Your taste is a personal thing, not an indicator of the health of the industry.
Not everything is bleak. The ending sequence for Frieren season 2 was entirely hand drawn with colored pencils[0] and the new Ghost in the Shell (coming out next month) is also hand drawn[1].
[0] https://www.youtube.com/watch?v=FY4Bx2qtkRM
[1] https://www.youtube.com/watch?v=nnRcKC4Rgsc
How much of this is actually drawn in Japan though? A lot of the drawing ends up being outsourced to places with cheaper labor like the Philippines.
Neat, I didn't know for Ghost in the Shell, I'll be watching it in few weeks then! Thanks for sharing.
PS: so weird to watch the trailer while working in XR. I never imagined as a kid I'd be programming in a headset but now it's banal. I even buy 2nd hand HMDs for 100 bucks. Weird times.
Noob Question: There is a famous parlor trick with generative networks(I think it was GANs but it might be some kind of diffusion based network.), you start with a canvas and draw a stick figure of what you want and the generative network draws the rest of it.
Do AI platform companies actually pre-train networks to do the same for hand drawn artists?
Related question: If they do train them to do that, are there any that train people for the "reverse": learn how to draw with paper and pencil by showing techniques only i.e only the "what" but not the "how" ?
re: parlor trick
Are you referring to:
https://github.com/lllyasviel/controlnet
?
Frieren is animated by Madhouse, though, which I feel like is one of the last bastions of high-quality Japanese animation.
Even if we're still far from the golden age, Dandadan (Science SARU; though I've heard the sheer effort bankrupted them) and the recent Witch Hat Atelier (Bug Films) are quite nice to behold.
Some modern romcoms like Kaguya-sama (A-1 Pictures) and The Dangers in My Heart (Shin-Ei) were nicely done too.
Things are obviously not going in the right direction, but the current accelerated fall in quality the consumer sees is more influenced by general unwillingness to spend and/or take risks from bean counting production committees.
https://archive.ph/aPMAr
Thanks!
After reading this article, I got curious whether there was a similar article from Japan, and there actually is. The average monthly income in the animation industry is 200,000 yen (about $1,300 USD), but the median working hours are 2,745 hours per year. That comes out to 225 hours per month, or 52 hours per week [1][2]. Considering that animators' work is essentially drawing labor, that's an insane amount of work. But even as total production costs and promotional effects grow, none of it reaches the workers on the ground. It seems like in modern industry, the value of promotion and fame outweighs what the laborers actually produce. Actually, when you think about it, this problem is happening across all sectors of society. Ultimately, it's a system where platforms intermediate and monopolize value.
Platforms concentrate their investment in IP and star creators, and the commercial success of these creators in turn increases the platform's value, creating a virtuous cycle. However, this success ultimately ends up concentrated among a small upper tier, while the vast majority are excluded.
The article essentially says the same thing.
It seems like we're in the age of platform capitalism. Come to think of it, the programming world feels similar too
[1]https://nafca.jp/news20241226/
[2]https://prtimes.jp/main/html/rd/p/000000004.000121993.html
There's a japanese docuseries called 'Manben' (available on youtube) that interviews a bunch of famous manga artists. Without fail they all talk about how much work, hours, suffering goes into the first years of making manga.
It's very interesting to watch and I highly recommend it. But it's also a GREAT advertisement for avoiding the industry.
I'm not sure that average across the entire industry is a good measure of how animators are paid for high budget, high talent features.
So basically they get paid peanuts and are overworked and so no one wants to do it.
Also seems like something AI could really cut into. You could have a master animator doing much of important thoughtful work and AI filling in the obvious as well as doing tweening (sound similar to programming)
Really, who needs a studio if you have the creative talent and ability to leverage AI for the grunt work. Or have a couple grunt work humans paid these rates to manage the tedious work of leveraging the AI to make it look seamless.
AI image generation is just not there yet. Say what you want about Luddites and whatnot, but the quality is just not good - the amount of effort given compensation that (especially Japanese) artists put into anime styled imagery makes too little sense that AI can't compete, even in generation time if time for retakes are accounted for.
You're not wrong. But there is a common perception that we value things made by humans more. The problem is that grunt work actually serves as a pipeline for industrial training. Even with AI, the distribution of value doesn't get resolved automatically.
Of course, I think it would be great if grunt work disappeared, but I believe skilled workers ultimately need grunt work. It's like saying that since AI automates everything, programmers don't need to know how to write methods. The core issue here is that grunt work, which AI excels at, plays an educational role in our society.
Of course, I admit my thinking is quite old-fashioned. This educational model could change. But I'm not sure whether that would be good in the long run. It could be beneficial in the long term. Humans evolve, after all.
I'll reserve judgment on that part.
> Of course, I think it would be great if grunt work disappeared, but I believe skilled workers ultimately need grunt work. It's like saying that since AI automates everything, programmers don't need to know how to write methods. The core issue here is that grunt work, which AI excels at, plays an educational role in our society.
It's not just educational. The more thinking you offload to AI, the more your own skills degrade [1] - and it makes sense, intuitively. If you repeat tasks, you gain experience and get good at it... but if you cease that repetition, eventually your skills break down.
[1] https://www.mdpi.com/2075-4698/15/1/6
I've thought a lot about how to characterize the difference between 3D and hand-drawn stuff. I think the core of it is:
- With hand drawn animation, you draw what you have in your head, first roughly, then you refine it
- With 3d animation, you first need to model everything, then rig it, then work with the bone system to get the motions you want or else mocap, and then set up rigs and stuff so that the mocap actors can do crazy movements, etc etc. Then maybe undo some of the scaffolding the 3d software does: disconnect bones, fake perspective. You have to fiddle with lighting, textures, etc. Or you don't, and just go with whatever's easy to do in the 3d software.
Which means that spontaneity and emotion, like I think this guy's arms should be all wiggly here, are lost. Yeah, you can hand animate then 3d animate on top of the hand animation, but in an industry that's using 3d to cut costs and not because it looks better, that's not going to happen (in any way that keeps the spontaneity).
3d is awesome in that once you've done a huge amount of up-front prep, the rest is easy to iterate on and tweak, but that's a large tradeoff.
I thought that this is one area where I think AI could be a force of good. Keep the animators doing the rough sketches, and AI comes up with the lines, handles the filling, and maybe adds colors with a guide. I haven't seen this yet.
---
I'm not sure I agree on the mentorship parts. IIUC all the major studios and famous animators weren't taught by someone. All the studios have unique flairs that they came up with just by playing around and copying Disney. And they got there without drawing hundreds of thousands of in-between frames for someone else.
I think that being taught the correct way to animate based on existing productions probably also reduces creativity in the field.
I wonder if the earlier creativity was due to voids though, and now that there's some amount of saturation it's harder to break in, or if somehow the increase in revenue from global interest somehow increases stakes and causes more downward pressure squashing out experimentation.
---
There's a lot I don't get about this article though. It says the demand is way up, but the treatment and pay for animators is terrible... why? I didn't see it addressed. Japan has a long history of "non-monetization" though, like refusing to sell digital music overseas, or regional restrictions on streaming content.
---
Lastly, I think there's still a lot of indie animation that gets glossed over. There are lots of independent animators making animations for music videos, for instance, or releasing small animations. I don't know if that grows into larger productions, but there's a level of creativity you'd never see if you just watch televised anime.
Interesting, the "training of new animators" mirrors what has/is happening in other industries.
When I started programming decades ago, an experienced programmer would review my work and help me out. That started ending in the very late 80s and 90s. By 2000 or so, you were on your own as a new employee. I even mentioned it to a high level manager a while ago, he said we expected people we hire to know what they are doing.
I have heard similar things have occurred in manufacturing too.
Do you think this is a global phenomenon or regional maybe just in the US or the Anglosphere?
Remember the joke about asking for 5 years of experience with a 5 month old framework?
This is part of the whole move fast and break things mantra. If you have to train people you aren't moving fast enough. And now they can bolt on AI turbochargers.
I feel like this might be the transition between an industry driven by apprentice-ship like guidance to an industry driven by credentialed training.
It'll be very interesting reading future studies on how this has negatively impacted entire generations. Hopefully people realize that you need to pay younger people living wages to learn skills, if they want people to have those skills in the future.
My (probably unpopular here) opinion is just the opposite. We need more of an apprenticeship model where you're not paid a bunch because you're still learning, and you probably bring negative-to-zero value initially. When a fresh-out-of-college junior engineer brings in SV style money, the expectation should be that they already know what they're doing.
In the trades you start off low pay because you're generally more in the way than helpful, but then you gain the experience and knowledge to be valuable.
Even the resident-doctor relationship is like this. Resident are overworked and poorly paid because they are more distracting than the value they add, then there's the big reward at the end.
The grad-student/professor model is kind of like this too except for all the pyramid scheme stuff that happens there.
I think most technical fields need to go to this model where the newbie commits to learning and trying to be valuable instead of rest-and-vest. And then once they're valuable they get paid more in proportion to the value they bring to the field.
In my small company we had to switch about 5 years ago to only hiring folks with lots of experience (10-15 years). We tried hiring younger fresh-out-of-college engineers, but "market rate" was too high and they required too much attention from senior staff and it made us unsustainably unproductive. We wanted to mentor and teach the next generation, but we couldn't afford it.
i love an apprenticeship model, i think most people would learn better and become competent at whatever position much quicker than straight book learning, for practically every job.
however, where you say: "My [...] opinion is just the opposite [...] where you're not paid a bunch ", are you saying the opposite of a living wage? how would you expect someone to, well, live during their apprenticeship? someone starving and worried about getting evicted or similar is not in a great head space to learn effectively.
Let in the chronicles be written, that widespread inability to read and write endeth the chronicles.. class of mistake. Let this be a lesson to all readers of chronicles.. the big failure starts three generations before the white paper appears..
> we expected people we hire to know what they are doing
I feel this is a generational thing. Many baby-boomer parents never took the time to teach their children any skill. They thought they learnt it by osmosis I guess. Their generation outsourced everything they could.
God forbid we invest in the future. Investors need those profits now!
We know where all the animators are, Mappa has them chained in the basement working on the next JJK season /s
Anime has never been more popular but honestly its quality has declined a lot. It's all isekai and supernatural-sexy-schoolgirl-whatever (kinda creepy) these days. Or the same old animes that go on forever.
Every year there's less and less animes that are worth the time to watch IMHO.
The same ratio of garbage to quality was available back then, there just wasn’t the same level of exposure to the range of output so the worst ones faded into history.
Funny, that was already true in 2006. It's the thing people keep saying, and yet anime keeps coming out.
Right now they're stuck in the whole "ten shows from the same budget, each run by their own 'committee' and each competing for cash that runs out well before the show's over, and the poor performance very quickly get less per episode". Great for networks, shit for shows. Even worse for animators who need to get paid a living wage.
A cursory glance of anime tv shows released in 2006 and there's a lot of good things I'd watch again: Honey & Clover, Kemonozume, Ergo Proxy, Code Geass, Haruhi Suzumiya, etc.
In 2026 the only thing that has captivated me, so far, was the Chainsaw Man movie and Dorohedoro S2.
Higurashi when they cry does the whole concept of time reset well (2006).
Trigun sequel is coming out this year.
2026 titles have caught on that trend of a really long specific title though.
“The Laid-Off Cheat-Granting Mage Enjoys a Second Lease on Life”
I think anime is a big enough category that every year there’s always at least a few good ones but it’s hard to get good signal to noise. But the whole tiny girl thing is 100% a trend that’s stuck around.
Akane Banashi was probably the best directed show this season.
You're into Chainsaw Man and Dorohedero but not JJK?
Maybe there are only two things you would watch but there are tons of others that a lot of people are watching. Your taste is a personal thing, not an indicator of the health of the industry.
What? This is a golden age of anime.
You could count the number of anime available in the west and worth watching 25 years ago on one hand, maybe two.
Now? If you can't find something, you're not looking hard enough. Off the top of my head, current anime that do not remotely fit your categories:
- Spy x Family
- Dandadan
- Dungeon Meshi
- Apocalypse Hotel
- Yumi no Tsugai (Daemons of the Shadow Realm)
- Kaiju No. 8
- Marriage Toxin
- Steel Ball Run
- The Summer Hikaru Died
- Akane-Banashi
- Dorohedoro
edit: formatting. edit 2: added Dorohedoro, good call down below
I did not know they made an anime of The Summer Hikaru Died. The manga was amazing, I need to check out how they adapted it.