DreamWorks Animation co-founder Jeffrey Katzenberg made himself unpopular in the animation industry when he predicted a few days ago that AI would make feature film production ten times more efficient and reduce the number of artists by 90%.
He said: “In the good old days, when I made an animated movie, it took 500 artists five years to make a world-class animated movie,” he said. “I think it won’t take 10% of that. Literally, I don’t think it will take 10% of that three years from now.”
Katzenberg, creator of the failed video app Quibi, is wrong on this prediction for at least four reasons:
1. AI will lead to a proliferation of second-guessing. Since everyone in the organization, including bean-counters and paper pushers, can generate alternate takes at the press of a button, it will be hard for any creative organization to commit to any set of decisions without facing the inevitable chorus of "Why don't we try this instead?" This effect will slow production, not speed it up.
2. Innovations won't come from the top down, but rather from the bottom up. AI techniques will surely lead to artistic innovations, new production methods, and even new art forms. But the animation production pipelines are so well established in studios like Disney, Pixar, and DreamWorks that they can't easily adapt to radically new toolsets. The filmmakers to watch are the little breakaway upstarts, the tiny mammals currently running around the ankles of the dinosaurs.
3. In practice they don't take you all the way to the finish line. Even though AI methods seem to magically produce immediate, sexy-looking results that would take teams of people weeks or months to produce, they still require the same kind of human expertise and effort to get really first rate results that the public will accept.
5 comments:
That last reason is really the only reason. The copyright office already ruled some time ago that AI generated art cannot be copyrighted, a big studio is not going to produce something for which they don't have total control over distribution.
It's odd how the people most eager to accept this technology tend to put forward a highly pragmatic worldview and dismiss artistic, creative and even ethical and human concerns, AND YET they seem to stake a lot on this imitative automation (openly advertised to heavily undercut requirements of qualified experience and labor) somehow resulting in the SAME consumer value as what's imitating, a notion contradicting basic pragmatic logic. I mean, why would they expect for the output 10% of the resources to justify anything less than a 90% reduction of the price of access to it?
New technologies always have a mix of good and bad. It's hard to tell where the balance will fall on this.
I imagine animators and artists working in animation would do well to look at the history of 3D animation and its effect on 2D animation jobs. 2D animators did not fare well in that transition and a wealth of animation knowledge was lost. The mass layoffs throughout the industry when Disney closed feature 2D production were sudden and painful. Some animators made the transition to 3D animation, but it is a different creative process more akin to stop-motion or very slow puppetry. Many senior 2D animators transitioned for a while and then decided to retire or transition into another field. Designers transitioned more smoothly and art departments expanded. 3D animation also opened up a wealth of more technical jobs. So, some good, and some bad.
AI animation tools also present a very different creative process with very little in common with previous methods. It is hard to predict who will transition well and who will not.
Bean counters have been interfering in productions since the beginning of film. But historically, production times have more often increased because directors wanted to make changes or try new directions. Bean counters tend to favor "good enough."
Bean counters are usually perfectly willing to... well, bean count. The potential of AI tools to save time seems more likely to affect quality than production time. A lot of current animation productions are automating in-betweens to cut costs. All of the new Netflix animated series are floaty, and the shows seem to be doing fine with audiences. Those are made with 3D tools, but they demonstrate a willingness among viewers to accept lower quality. Cutting animation quality isn't anything new. The Simpsons and South Park proved a long time ago you could shortcut animation quality and still make a successful and entertaining show. Both shows used relatively small animation staffs. And bean counters know this. The bulk of current animation production is straight-to-streaming, not feature.
Studio lawyers aren't trying to remove AI tools from all areas of the production pipeline. They are actively trying to extend AI tools in some areas. That was the whole point of the writers and actors objections to AI and other automation. And the agreements did not resolve this issue. The opt-in/opt-out option seems the most likely way forward. Motion capture and crowd animation have been around a long time and have set many de facto copyright precedents, so the end result is likely to favor studio interests where they own the motion capture data.
This is not significantly different from studio-owned designs and models. There is currently nothing to stop a studio from using AI tools, or any other tools, to create new intellectual property based on intellectual property it already owns.
Wes, thanks for those perceptive insider insights. I'm curious: which executives at Disney made the decision to gut 2D animators in favor of Pixar-lite 3D? Was Katzenberg at Disney then? And how are the Mouse-House execs now responding to the innovations by Fortiche (Arcane) and Sony Pictures Animation (Spider-Verse)?
I think you're right to see AI as a new tech with both good and bad potentials. Who knows how the public will respond to new work that bears the AI look? (Witness the response to the "Secret Invasion" intro.) And there are other factors outside of this tech that comes into play, such as allowing commercial products drive content, such as the Lego and Barbie movies and the inevitable attempts to repeat those successes.
James, I'm more an industry bystander than a true insider. So, take my observations with a grain of salt. I'm sure people like Aaron Blaise have a much more accurate perception about this than I do. My knowledge is mostly second-hand.
At the time Disney closed its 2D feature animation, I knew a lot of animators (both 3D and 2D), concept artists, and designers through a high school animation mentorship program (I was teaching high school at the time), and through the American Animation Institute (where I was taking life drawing, panting, animation design, and animation classes). My previous experience was as a set designer bouncing between theme parks, film, television, and theater. I also had some experience, quite accidentally, proofreading film rights contracts for authors during the loosening of the manager-agent rules.
I hope the legal and professional environment evolves ahead of the technology. I think there needs to be some reasonable protection for humans to continue doing human things for a living. The AI jobless utopia doesn't strike me as realistic, or desirable. At the same time, I hope we don't throwaway benefits of the technology because of fear.
I'm not sure how Disney is viewing the new 2D feature animations today. I like to think with chagrin. Then again Bob Iger's back (see below). But I think it is fantastic that Sony is embracing 2D animation--and in a particularly fresh way. It is a sign of how things often turn around in unexpected ways. At the time Disney closed its 2D feature animations, Sony was like a cult of 3D. Those guys were true believers.
Katzenberg was at Dreamworks when Disney closed feature 2D animation. He oversaw the end of their feature 2D animation in the late 1990s and early 2000s, long before most of the other studios. At the time, Pixar was the only feature animation company turning a profit. There were three common beliefs: audiences preferred 3D animation, 3D animation allowed for a more industry-standard story development process (closer to live-action films), and 3D assets would be easier to use on subsequent films (which would potentially save time).
From what animators told me, Katzenberg handled the transition fairly well. And 2D animators who wanted to remain in 2D still had opportunities at other studios. I had the opportunity to tour Dreamworks when they were working on Prince of Egypt, and later when they were working on Kung Fu Panda. The art department had grown and taken over the 2D animation building (nothing nefarious, 3D animation doesn't play well with big windows and natural light). 3D animation was in completely dark windowless rooms, which I can understand not being appealing to a former 2D animator. But my understanding is that Dreamworks went out of its way to provide training for 2D animators who wanted to make the transition to 3D.
Bob Iger was in charge of Disney when they cut 2D feature animation. The massive layoffs started after the 2008 housing crash, but the layoffs were uneven. Layoffs for animators working on Princess and the Frog were delayed, but one of the animators I took life drawing with told me it was depressing to watch them repurpose the feature animation building for 3D. They were hiring new 3D animators during the layoffs to take over Tangled and to work on Frozen. 2D had moved into a warehouse building. I remember a month during this time when Disney sold or gave away hundreds of animation desks. If someone could get you on the lot with a truck you could drive away with one.
Post a Comment