Sony Turns to AI to Tackle Rising Animation Costs and Animator Shortage | Hacker News Reader

Sony Turns to AI to Tackle Rising Animation Costs and Animator Shortage

cartoonbrew.com

28 pointsDevScout46 comments

There is no animator shortage. They just don’t want to pay up.

Baumol cost disease

Cost reduction is a major focus

How long does it take to learn? Are other jobs that draw from the same talent pool becoming more efficient?

As someone learning 3D animation for indie games, it seems to take at least a year to get good enough. And even after spending that much time drinking from the firehose of information that Youtube is, I suspect that someone with a 4-year degree in animation would have a significant leg up over someone learning by themselves.

I think there are a lot of people who are going to school for animation, but maybe supply and demand are mismatched? But at the same time, a lot of VFX, animation and game studios are doing layoffs, so there probably shouldn't be that much a shortage of talent? I suspect GP comment may be right that the studios that are not able to find animation talent may just not be willing to compete on salaries.

Anyone can pick up a pen and draw. But animation/cartoon/anime are both a product and art, and require time for those who draw to grow from amateur to master. Some will get there in no time, and some will take time to get there. There are no one size fit all equation.

lol "Tackle Rising Animation Costs"

There really is an animator shortage in anime. There's a popular blog with multiple articles on it: https://blog.sakugabooru.com/2022/08/05/the-layout-crisis-th...

Though simply paying more sounds like it would fix the issue.

They are suggesting the shortage exists because nobody is getting paid well enough to continue onwards or join in. These folks are also necessary to raise the next generation of animators.

It's not just that, it's also that training opportunities have become rare.

Expecting to hire employees that are 100% trained to do the specific vacancy on day 1 is why the market is shit right now.

Yeah the shortage is because the pay is shit and the companies don't want to have any training burden. Self-created shortage.

I think plenty of companies would like to spend time training but aren't able to because the budget is too small. Still makes Sony reasonable for the shortage as you say.

If your budget is too small for training its too small for running a business outside of 1 person.

It's 100% this.

To say this is to argue there is never a shortage of anything.

Its the same thing as people claiming there’s an “accountant shortage”. Both are nonsense. There’s a shortage of people who want to do the job for pennies

Same with tech worker “shortages” in places like the UK and Australia.

No, what you have is a shortage of people willing to fill your vacancy at those prices.

Particularly in Perth there is a whole shadow economy of people working remotely for overseas firms, because local positions don’t pay the bills. This doesn’t say “shortage” to me, it says “cheap”.

And in the UK most small/medium businesses seem stuck in a vicious cycle of low pay, low productivity and low expectations.

Hayao Miyazaki's thoughts on an artificial intelligence.

https://www.youtube.com/watch?v=ngZ0K3lWKRc&t=47s

To be fair, the footage they showed him was just not good. They tried to play it off as as horror genre clip, but knowing how bad "AI" is/was at the time makes me think that wasn't the original intent. That was uploaded 7 years ago. GenAI has made improvements since, but it still doesn't dispel his point

There is also a huge shortage of BMWs in the $5000-$7000 purchase price range.

That's just a local problém, there's no such shortage in Europe. Guess I'd just need a really cheap shipping option to the US to make some profits ;)

>> That's just a local problém, there's no such shortage in Europe.

Your point is valid even in the context of my humor actually. Many of these problems are highly local.

For example, the "shortage" of things in NYC are usually because people dont account for local constraints. My colleagues in NYC complain about a "shortage" of nannies, yet cant seem to see that the living wage in NYC is super-high. Why would a nanny work for below-living wages? I'm sure that with commensurate income, the shortage would disappear.

That's right, straight to the point.

> “Whether it’s for games, films or anime, we don’t have that much IP that we fostered from the beginning,” the company’s chief financial officer Hiroki Totoki told FT. “We’re lacking the early phase [of IP] and that’s an issue for us.”

I read this from Playstation before, but the quote makes 1% more sense when it's expanded to anime and films.

It's still stupid because they are doing the exact same thing in film as in games these days: relying on Spiderman (among other 3rd party franchises) to basically carry their portfolios with some of the best talent in the business. It seems like their narrative doesn't match their actions.

(I won't be too harsh on the anime sector, because being realistic: 90+% of anime are adaptations of something else and manga/ln properties do tend to actually stay in ownership of the author).

>with costs ballooning 40-60% due to growing budgets in Japan and a limited number of animators in the anime industry.

One look at the budget of an anime and the compensation of an animator and these claims instantly shift to "animators finally getting tired of being abused". Even the games industry can't compared to how utterly grinded the animators are, while paid even worse to boot. That's a very deep topic to go into itself so I'll just say that the structure of the anime committee needed to die yesterday. It's a borderline cabal.

Someone wanting a deeper dive into the structure can watch this video as an introduction: https://www.youtube.com/watch?v=iExwO1v_V-s

This is great news! I look forward to seeing new Sony animated movies with the human characters having eyes on the sides of their heads, 8 fingers on their hands, and 3 legs.

It's not like a lot of animated characters are realistically human anyway.

They're realistically humanoid (two arms, two legs, head with 2 forward-facing eyes and mouth), which is more than I can say for a lot of AI-generated "art" I've seen.

They are unrealistic in a consistently stylized way, unlike AI-generated images which are simply broken.

I remember animating things years ago when I used Macromedia Shockwave. One of the first things I learned about was tweening, where you had a start frame and end frame and the software just animated the gaps for you so your animation worked. If there were software that animated whole frames with this in betweening process, I foresee that becoming the new way to do things. Hopefully software like this makes its way into open source tooling where others can take advantage of it. Maybe it even already exists by now. :)

tweening is indeed a fundamental concept in animation, and it's great to see you're nostalgic about the early days of Macromedia Shockwave.

I think you're onto something with the idea of software that can automatically animate entire frames with tweening in between. This could potentially revolutionize the way we create animations, especially for tasks like data visualization, scientific simulations, or even video game cinematics.

There are already some tools that can generate animations using tweening, such as Blender's animation tools or even some proprietary software like Adobe After Effects. However, I'm not aware of any open-source tools that can do this on a large scale.

That being said, there are some interesting projects like OpenToonz, which is an open-source 2D animation software that uses a tweening-based approach. It's not exactly what you're looking for, but it's a good starting point.

I'd love to see more open-source tools that can automate the animation process using tweening. Do you think this could be achieved using machine learning algorithms, or would it require a more traditional programming approach?

Also, have you explored any of the newer animation tools like Houdini or Nuke, which have some impressive animation capabilities? Maybe we can learn from their approaches and apply them to open-source tools

If you take a look at crunchyroll today you'll see this push already happening. A flood of low quality light novel adaptations has appeared and is seemingly overtaking their front page. Light novels because there are so many now, and in one place for producers to sort by popularity rather than quality. The animation is often poor and I won't be surprised if it's already the result of assistive genai tech. The combined factors must make it simpler for then to try and see what works.

So will they be able to copyright their output?

Why wouldnt they?

I mean theres been a few articles on this very site that dispel the nonsense hyped up over the Thaler case.

Its really become a meme at this point to fud about AI copyright.

Are people actually going to pay to watch AI slop though?

It isn't made for humans anymore.

Yes, but people would probably not realize that it was made with significant assistance from AI.

I see this as one of the few legit use cases for current tech so far. The animators can draw key cels and the ai can do inbetweens.

Isn't drawing inbetweens one of the ways a lot of animators break into the industry? Key animation is usually an expert position for lead animators, but how will anyone become a key animator without being able to learn the process?

I would think the software with AI becomes available to the students for learning and starting with key animations becomes the thing you learn how to do as a novice. I don’t know how feasible that is, but in my experience software and techniques become more obtainable over time.

Who does that style animation today? It seems so much of it is 3D animated with a cell style render. I'm not an anime fan, so maybe somebody is still doing cell based animation, but that would blow my mind. So breaking into the industry seems very strange to me.

The problem of denying animators necessary fundamental experiences by automating the opportunities to learn that experience away with AI is the same problem either way. You can't just arbitrarily remove human skill and effort from the creative process and expect the end result to be more efficient and of equal or greater quality, anymore than you can expect the quality of a codebase to always improve by removing lines of code. The problem here isn't explicitly that inbetweeners exist, it's that animators aren't being paid commensurate to their Abilities.

AI isn't simply a tool like Photoshop or Blender, which augment an artist's existing skill. AI is designed to simulate and replace human creativity in the same way automation replaced physical human labor elsewhere. That's a fundamental difference. Replace physical animation with cel animation and the skill and talent of the animators still matter, even translated to another paradigm. Replace a human artist with AI, and the value of their talent is just gone. Even if the AI is trained on that specific artist's work, it can't grow and expand its horizons the way that human artist can, it can only iterate on what data the model already makes available, and express generalities based on that.

And why bother have human beings involved at all when you can just train an AI to do the whole thing? The whole business case for AI is devaluing the market value of human creative/knowledge based labor or automating it away altogether.

I never suggested using AI. You've confused the topic. I simply asked who was animated with hand drawn cells. Switching to a 3D animated but cell rendered output has nothing to do with AI. Just look at Beauty and the Beast. Look at Archer. Look at so many anime titles. Hand drawn/painted cells was just not practical, and the whole industry changed.

None of that suggests using AI.

Anime is traditionally done frame by frame. Mostly America focused on the 3D craze or puppets/Flash-style.

Traditionally doesn't mean squat for someone trying to break into the industry today as the original comment proposed. Mostly America is also a bizarre comment as well. Just look at all of the animated content from Disney, WB, etc to see all of the hand drawn/painted frame-by-frame. So I'm really not sure of your point

If I stubbed out, for example, a Java or c++ class and gave “an ai” a well-crafted prompt to write the guts, it would go like this: either spend an absurd about of time writing and rewriting prompts to get a garbage result that needs re-writing/debugging, or spend the same amount of time writing the code myself.

Edit: I forgot the important bit: why is animation any different.

Reply on news.ycombinator.com