How much of this is AI automating? That drawing of the car design, the storyboarding... isn't generative AI aiming to do away with most of it?
I've been seeing some of the lengthier AI films on X (i.e. the one about badgers vs. rats that looks like Lord of the Rings), and there is certainly a lot of work put into them. However the type of work that goes into them is substantially different than what is shown in this article about Zootopia. There are a lot less decisions made by humans. AI is driving a lot of the creative decisions. In the end you get something good, but is it still riding on the decisions of the training data that came from humans like the ones who made Zootopia by hand crafting every element of it? I think this is an interesting question.
Of course, but then you're outsourcing it to the vision of the AI model. I'm just wondering how much of this gap is being overlooked. There is definitely some control being lost but we still don't know how much or how much it's really worth. Will we just get the same types of films with AI over and over? I think it's early but there are some clues by seeing how much work was put into "traditional" workflows that have control over every single detail.
> is it still riding on the decisions of the training data that came from humans like the ones who made Zootopia by hand crafting every element of it?
How could it not be? Even if AI spews out a bunch of stuff that humans end up tweaking this limits future human creative output to modifications of whatever AI has already been trained on. It can only lead to a homogenization of art as AI regurgitates past works over and over again, only to be further distilled as new models are trained on little else but it's own slop or variations on it.
What makes animation interesting are the creative decisions. Character design. Prop design. Background design. Story. The more companies choose to use AI instead of letting artists make those creative choices, the more derivative and less interesting animation will become.
I remember seeing a Pixar documentary a really long time ago where someone talked about attention to detail. The designers would fully texture out the inside of a dresser draw that will never be shown in the film. The idea was imagine how much attention to detail and time is spent on the things that will be seen. Pure craftsmanship.
> The designers would fully texture out the inside of a dresser draw that will never be showing in the film. The idea was imagine how much attention to detail
Hehe, as an ex FX animator/TD, this sounds kinda funny to me, there’s likely more to the story. If they knew for sure the textures weren’t going to be shown, it’s a waste of time, not craftsmanship or attention to detail. But it’s plausible because often modeling & surfacing departments have to start work before the shots & camera are fully planned. And sometimes there are camera changes late in the production. It could be that someone’s shot that saw the inside of the dresser drawers got cut, that surfacing expected the drawers to be open in a shot, or that a scene inside the drawer was in storyboards but didn’t make it to layout, stuff like that happens a lot. But no productions are having artists work on stuff that they know will never be shown in the film.
That's interesting. So, filling in details on demand with something like Gaussian splats and then backfilling them later on could save a lot of "wasted" time.
You're thinking of Steve Jobs and his attention to detail on the inside of Apple computers; a trait he learned from his father as detailed in the book about him by Walter Isaacson
Quite the opposite. Texturing something they know will never be seen was (at least in those days) time not spent paying attention to detail on the things that will be.
I do understand it from an artistic perspective, and who knows when and how an asset will be used in the future; but the premise doesn't appear to hold water under scrutiny.
Now go watch "Kono Express". Disney laid off an animator from the Star Wars factory, and he's cranking out stuff about as good as Disney Animation all by himself. With Unreal Engine and AI tools, of course.
It's been a long time since there was an ink and paint department.
I opened this with an open mind, but it felt dead and AI immediately to me. Neural Viz, however, had succeeded in making things I find watchable all by himself[0] by working around some of the current weaknesses and by having stellar human writing. Perhaps unsurprisingly, it seems he got burnt out from doing it all alone so it's been a while since the last upload. The Remo Green one is my favorite[1].
Neural Viz had really good consistency and really good scripts, also importantly knew to write the story around the limitations in what AI could do, including using non-human characters to avoid that uncanny valley that I immediately see in the Kono Express video. Truly a master at his craft, it's a shame he hung it up!
Some people are making absolute banger content with AI. It's all in who's at the helm and what their storytelling capabilities and taste/aesthetics are.
I think another important point that AI is revealing is how diverse the population's preferences are. Legacy entertainment is nowhere as precisely tailored, because it tries to appeal to a wide audience.
Personally, all the Harry Potter AI shorts recently have been rather amusing, although they also vary significantly in detail and wit.
> Legacy entertainment is nowhere as precisely tailored
In some instances, it's even engineered to be broadly appealing. See Disney and Marvel.
On the other end of the spectrum, we have niche-specific creators who appeal to specific genre fans and can package their own tone and taste. That's actually really cool.
It's kind of like what happened to music. Electronic music, the DAW, and home recording opened the avenue for indie artists and all of the various genres.
No, even one of the human worst human slop is still more enjoyable than whatever polished AI slop. In China the Niu Lai Film is way more successful than many AI slop.
I haven't watched content for a long long time before AI slop started appearing, because the topics and the stories were boring and repackaged thirty times and ultimately very safe (need to be profitable).
AI has enabled small creators with new ideas and people with nothing but love of storytelling to be competitive with corporate slop.
Yes most of it's bad, but some of it is very innovative. No Hollywood slop has been innovative for a long time.
Around 2010–2012 I made videos for Vocaloid songs, with the art coming from separate illustrators. Many of them wouldn't send their illustrations split into layers, which made animating them a pain. And even though it was a music video and I asked for more drawings, some just wouldn't listen and sent a single piece with way too much effort poured into it.
Dreamworks used Renderman(Pixar’s Renderman) along with several other commercial package for a Shark Tale. When Dreamworks acquired PDI, who made Shrek under contract for DWA, there was essentially a war between the two pipelines and PDI won. I was on the PDI side of the company, but there was a lot of positive things about the Renderman pipeline.
Hey I was on the PDI side as well, ahoy, nice to see you here! I had forgotten about Shark Tale using RenderMan, but I will say to the degree there was competition and to the degree that PDI won any pipeline battles, DWA eventually won the war. :P ;) There was plenty of cooperation as well, it just takes time for two companies to integrate, and plenty of people on both sides were quite attached to the cool tools they made. (I was certainly guilty!)
PDI’s toolset was pure magic for it’s time. When I joined, I was in heaven with how easy it was to make procedural animation, build new effects that crossed pipeline stages, render anything you could imagine… I wished I could share it with researchers and hobbyists.
We all had to imagine making our next show using each others pipeline and that was fun. In the end, I think Richard, Jono and other PDI folks ended up having a bit more influence because Shrek was so darn successful.
Custom vs. commercial was a constant issue. Were you around for the whole NextGen effort? DWA spent a fortune on some custom tools that only were used on a show, or maybe just a few sequences, before they were replaced with commercial tools.
Oh definitely - custom vs commercial. I remember the R&D team having to write a yearly report to justify the development expenditure, explain repeatedly why we were writing custom tools & renderer when these things were available off the shelf. I heard a few strong rants here and there about how the suits just didn’t understand. I remember the term “NextGen” being thrown around but I don’t remember if I lived through it or it was just getting started when I left (before S3). Jono I and tried for a while to stop having our shaders rewritten in C++ (without otherwise changing them nor adding any features). We eventually gave up and just kept building new things, but that was a fairly successful strategy to nudge the IP, tech, and support southward. Possibly the right strategy in retrospect, but felt a little rough at the time.
“Better” is a loaded word. When you write your own renderer, you have a lot more control and a lot more support. Renderers at the time had not all converged on the same algorithms, so you could have a competitive advantage by writing a good one. PDI’s shader system was different than RenderMan’s and had some real advantages — compiled shaders, geometry & light shaders (not just materials), particles & particle shadowing, compositor integration, etc.. The renderer also needs to integrate into the pipeline deeply, so PDI’s renderer fits the PDI/DW pipeline better, whereas with RenderMan you have to try to mold the pipeline to fit the renderer in.
It wasn’t the renderer as much as the pipeline. The renderer is just one process in a stunning amount of complexity that must be managed. We may even agree that Renderman was better than the PDI renderer but there was so much much more.
Because all this animation tech is a moving target, each production's management chooses the software and versions of that software a production will use and then freeze it. So the studios already have in place a subsystem for every production to be running different software for different productions, doing the same jobs as other productions using other software. This allows some experimental renderer to be used in some sequences, and so on, as well as trials of all kinds of software too, such as AI tracking of camera paths, and the soft and hard objects in their view, AI assisted rendering, assisted color timing assistance and then experiments directly with AI video models in a hybrid pipeline.
Both companies each had their own renderer & rendering development team before the acquisition. And their own pipelines. A couple decades later, that’s still true.
This reminds me of a very similar Disney exhibition at Miraikan in Tokyo back in 2017 that showed what went into making an animated film. It was a fascinating exhibition, that focused a lot on Moana but had a range of other films on show that they explained the process behind. The fact we have so many people dedicated to one short movie is amazing and in my eyes always worth the effort.
> The fact we have so many people dedicated to one short movie is amazing and in my eyes always worth the effort.
i’m convinced this is where much of the magic comes from.
it’s the same as why often music written and played by an group of musicians who specialize in their instrument is often so much better than the music from a solo who records each instrument themselves.
Everything is. Including this comment. And yours. But if that's the thought upon looking at that post that you need to share, you might need to touch grass, as they say.
The earliest mention I could find for the page is from Feb 2022, which predates ChatGPT, so it wasn't made for that purpose.
Very relevant, here is an animated film I just worked on.
Coming from software development, I had to learn everthing on my own and studied Disney and Pixar quiet a bit, so this was very helpful.
It was very fun and hard to learn, would love your thoughts :) https://www.youtube.com/watch?v=XcZ-nGvkg0U
How much of this is AI automating? That drawing of the car design, the storyboarding... isn't generative AI aiming to do away with most of it?
I've been seeing some of the lengthier AI films on X (i.e. the one about badgers vs. rats that looks like Lord of the Rings), and there is certainly a lot of work put into them. However the type of work that goes into them is substantially different than what is shown in this article about Zootopia. There are a lot less decisions made by humans. AI is driving a lot of the creative decisions. In the end you get something good, but is it still riding on the decisions of the training data that came from humans like the ones who made Zootopia by hand crafting every element of it? I think this is an interesting question.
The point of storyboarding is to determine what you want to make. I guess you can consider it part of the "prompt."
Not something that can be automated away assuming you still have preferences at all.
Of course, but then you're outsourcing it to the vision of the AI model. I'm just wondering how much of this gap is being overlooked. There is definitely some control being lost but we still don't know how much or how much it's really worth. Will we just get the same types of films with AI over and over? I think it's early but there are some clues by seeing how much work was put into "traditional" workflows that have control over every single detail.
> is it still riding on the decisions of the training data that came from humans like the ones who made Zootopia by hand crafting every element of it?
How could it not be? Even if AI spews out a bunch of stuff that humans end up tweaking this limits future human creative output to modifications of whatever AI has already been trained on. It can only lead to a homogenization of art as AI regurgitates past works over and over again, only to be further distilled as new models are trained on little else but it's own slop or variations on it.
What makes animation interesting are the creative decisions. Character design. Prop design. Background design. Story. The more companies choose to use AI instead of letting artists make those creative choices, the more derivative and less interesting animation will become.
I remember seeing a Pixar documentary a really long time ago where someone talked about attention to detail. The designers would fully texture out the inside of a dresser draw that will never be shown in the film. The idea was imagine how much attention to detail and time is spent on the things that will be seen. Pure craftsmanship.
> The designers would fully texture out the inside of a dresser draw that will never be showing in the film. The idea was imagine how much attention to detail
Hehe, as an ex FX animator/TD, this sounds kinda funny to me, there’s likely more to the story. If they knew for sure the textures weren’t going to be shown, it’s a waste of time, not craftsmanship or attention to detail. But it’s plausible because often modeling & surfacing departments have to start work before the shots & camera are fully planned. And sometimes there are camera changes late in the production. It could be that someone’s shot that saw the inside of the dresser drawers got cut, that surfacing expected the drawers to be open in a shot, or that a scene inside the drawer was in storyboards but didn’t make it to layout, stuff like that happens a lot. But no productions are having artists work on stuff that they know will never be shown in the film.
That's what I'd have thought too, it's more of a just in case thing, if the director later wanted a shot inside the drawer.
That's interesting. So, filling in details on demand with something like Gaussian splats and then backfilling them later on could save a lot of "wasted" time.
You're thinking of Steve Jobs and his attention to detail on the inside of Apple computers; a trait he learned from his father as detailed in the book about him by Walter Isaacson
Jobs funded and bought Pixar from George Lucas so it's no surprise the same level of detail went into both his companies.
Quite the opposite. Texturing something they know will never be seen was (at least in those days) time not spent paying attention to detail on the things that will be.
I do understand it from an artistic perspective, and who knows when and how an asset will be used in the future; but the premise doesn't appear to hold water under scrutiny.
Now go watch "Kono Express". Disney laid off an animator from the Star Wars factory, and he's cranking out stuff about as good as Disney Animation all by himself. With Unreal Engine and AI tools, of course.
It's been a long time since there was an ink and paint department.
I opened this with an open mind, but it felt dead and AI immediately to me. Neural Viz, however, had succeeded in making things I find watchable all by himself[0] by working around some of the current weaknesses and by having stellar human writing. Perhaps unsurprisingly, it seems he got burnt out from doing it all alone so it's been a while since the last upload. The Remo Green one is my favorite[1].
[0]https://www.wired.com/story/the-future-of-ai-media-parody-of...
[1]https://www.youtube.com/watch?v=5bYA2Rv2CQ8
> I opened this with an open mind, but it felt dead and AI immediately to me.
Episode 1 is. He's getting better at this. By episode 3, the characters have come alive and the scenery is really good.
Neural Viz had really good consistency and really good scripts, also importantly knew to write the story around the limitations in what AI could do, including using non-human characters to avoid that uncanny valley that I immediately see in the Kono Express video. Truly a master at his craft, it's a shame he hung it up!
[flagged]
> he's cranking out stuff about as good as Disney Animation all by himself
https://www.youtube.com/watch?v=ikm5lP0zsi0
Yeah... no. Not at all. It's just classic AI slop.
The fact that someone can honestly say this is anywhere near the level of any professional animation is truly baffling.
Just the first two shots alone are full of horrible, stilted and morphing animation, no consistency between shots, no particular art style at all.
This is not a creative work or art in any form I recognise.
> classic AI slop
A lot of AI content is better than human slop.
Gossip Goblin is fantastic.
Some people are making absolute banger content with AI. It's all in who's at the helm and what their storytelling capabilities and taste/aesthetics are.
I think another important point that AI is revealing is how diverse the population's preferences are. Legacy entertainment is nowhere as precisely tailored, because it tries to appeal to a wide audience.
Personally, all the Harry Potter AI shorts recently have been rather amusing, although they also vary significantly in detail and wit.
> Legacy entertainment is nowhere as precisely tailored
In some instances, it's even engineered to be broadly appealing. See Disney and Marvel.
On the other end of the spectrum, we have niche-specific creators who appeal to specific genre fans and can package their own tone and taste. That's actually really cool.
It's kind of like what happened to music. Electronic music, the DAW, and home recording opened the avenue for indie artists and all of the various genres.
>A lot of AI content is better than human slop.
No, even one of the human worst human slop is still more enjoyable than whatever polished AI slop. In China the Niu Lai Film is way more successful than many AI slop.
That is completly your opionion,
I haven't watched content for a long long time before AI slop started appearing, because the topics and the stories were boring and repackaged thirty times and ultimately very safe (need to be profitable).
AI has enabled small creators with new ideas and people with nothing but love of storytelling to be competitive with corporate slop.
Yes most of it's bad, but some of it is very innovative. No Hollywood slop has been innovative for a long time.
>That is completly your opionion.
No, this is not only my opinion. Like i said, that human slop movie is way more successful than many polished AI movie in China.
So are there more such articles “what goes into making <X>”? that you would recommend? - eg- games, physical objects, hardware, etc?
Around 2010–2012 I made videos for Vocaloid songs, with the art coming from separate illustrators. Many of them wouldn't send their illustrations split into layers, which made animating them a pain. And even though it was a music video and I asked for more drawings, some just wouldn't listen and sent a single piece with way too much effort poured into it.
Hm that’s interesting, I would have thought that Pixar and Disney Animation use the same render software, but no.
Dreamworks used Renderman(Pixar’s Renderman) along with several other commercial package for a Shark Tale. When Dreamworks acquired PDI, who made Shrek under contract for DWA, there was essentially a war between the two pipelines and PDI won. I was on the PDI side of the company, but there was a lot of positive things about the Renderman pipeline.
Hey I was on the PDI side as well, ahoy, nice to see you here! I had forgotten about Shark Tale using RenderMan, but I will say to the degree there was competition and to the degree that PDI won any pipeline battles, DWA eventually won the war. :P ;) There was plenty of cooperation as well, it just takes time for two companies to integrate, and plenty of people on both sides were quite attached to the cool tools they made. (I was certainly guilty!)
PDI’s toolset was pure magic for it’s time. When I joined, I was in heaven with how easy it was to make procedural animation, build new effects that crossed pipeline stages, render anything you could imagine… I wished I could share it with researchers and hobbyists.
We all had to imagine making our next show using each others pipeline and that was fun. In the end, I think Richard, Jono and other PDI folks ended up having a bit more influence because Shrek was so darn successful.
Custom vs. commercial was a constant issue. Were you around for the whole NextGen effort? DWA spent a fortune on some custom tools that only were used on a show, or maybe just a few sequences, before they were replaced with commercial tools.
Oh definitely - custom vs commercial. I remember the R&D team having to write a yearly report to justify the development expenditure, explain repeatedly why we were writing custom tools & renderer when these things were available off the shelf. I heard a few strong rants here and there about how the suits just didn’t understand. I remember the term “NextGen” being thrown around but I don’t remember if I lived through it or it was just getting started when I left (before S3). Jono I and tried for a while to stop having our shaders rewritten in C++ (without otherwise changing them nor adding any features). We eventually gave up and just kept building new things, but that was a fairly successful strategy to nudge the IP, tech, and support southward. Possibly the right strategy in retrospect, but felt a little rough at the time.
I wonder why PDI was better than Renderman ?
“Better” is a loaded word. When you write your own renderer, you have a lot more control and a lot more support. Renderers at the time had not all converged on the same algorithms, so you could have a competitive advantage by writing a good one. PDI’s shader system was different than RenderMan’s and had some real advantages — compiled shaders, geometry & light shaders (not just materials), particles & particle shadowing, compositor integration, etc.. The renderer also needs to integrate into the pipeline deeply, so PDI’s renderer fits the PDI/DW pipeline better, whereas with RenderMan you have to try to mold the pipeline to fit the renderer in.
It wasn’t the renderer as much as the pipeline. The renderer is just one process in a stunning amount of complexity that must be managed. We may even agree that Renderman was better than the PDI renderer but there was so much much more.
Because all this animation tech is a moving target, each production's management chooses the software and versions of that software a production will use and then freeze it. So the studios already have in place a subsystem for every production to be running different software for different productions, doing the same jobs as other productions using other software. This allows some experimental renderer to be used in some sequences, and so on, as well as trials of all kinds of software too, such as AI tracking of camera paths, and the soft and hard objects in their view, AI assisted rendering, assisted color timing assistance and then experiments directly with AI video models in a hybrid pipeline.
Both companies each had their own renderer & rendering development team before the acquisition. And their own pipelines. A couple decades later, that’s still true.
Popular in:
2022 (276 points, 167 comments) https://news.ycombinator.com/item?id=30241966
2024 (118 points, 22 comments) https://news.ycombinator.com/item?id=41204228
Oh, CGI, not cell animation, like in an animated film.
Previous discussion: 09.feb.2022 https://news.ycombinator.com/item?id=30241966 166 comments
This is exactly how I break down my AI films, but with a bit more primitives!
what primitives do you use?
You should probably break them down a little further and throw them straight away in the bin.
AI films will be as normalized in 20 years as CGI is right now, even though both had their controversies when they started out.
This reminds me of a very similar Disney exhibition at Miraikan in Tokyo back in 2017 that showed what went into making an animated film. It was a fascinating exhibition, that focused a lot on Moana but had a range of other films on show that they explained the process behind. The fact we have so many people dedicated to one short movie is amazing and in my eyes always worth the effort.
> The fact we have so many people dedicated to one short movie is amazing and in my eyes always worth the effort.
i’m convinced this is where much of the magic comes from.
it’s the same as why often music written and played by an group of musicians who specialize in their instrument is often so much better than the music from a solo who records each instrument themselves.
another place to go for an animated film
https://www.cbc.ca/news/entertainment/wildwood-review-laika-...
https://www.laika.com/our-films
Thanks for the links, will check them out.
[dead]
> Each frame is on screen for approximately 0.04 seconds
Is this training material for LLMs?
This comment (c) copyright by amelius, not for training AI.
Everything is. Including this comment. And yours. But if that's the thought upon looking at that post that you need to share, you might need to touch grass, as they say.
The earliest mention I could find for the page is from Feb 2022, which predates ChatGPT, so it wasn't made for that purpose.
https://www.reddit.com/r/Screenwriting/comments/slfb1q/disne...
Thanks, I updated the license on my comment.
License doesn't matter. They're going to take it. Copyright is dead (if you're an AI company)
Reddit's Terms of Service allow for AI training of any material you post here, and you agreed to that before posting your comment.
This isn’t Reddit?
Very relevant, here is an animated film I just worked on. Coming from software development, I had to learn everthing on my own and studied Disney and Pixar quiet a bit, so this was very helpful. It was very fun and hard to learn, would love your thoughts :) https://www.youtube.com/watch?v=XcZ-nGvkg0U