It’s been just over a year since ChatGPT launched, kicking off widespread discourse about the potential of artificial intelligence. Some kids producers have greeted the tech with skepticism and trepidation, while others are viewing it as an indispensable tool for producing animation faster and cheaper—a godsend for an industry facing deep financing cuts and rising costs.

Historically, animation studios have had to sacrifice quality for speed and scale. But with AI, it’s possible to have both. And that may be exactly what’s required to keep up with the ever-increasing need for more premium content everywhere that kids want it.

Demand in the animation industry for generative AI tools (which can initiate text, images and content) is exploding, with more and more companies adopting them or creating their own. And the market value of generative AI in animation is expected to grow from US$900 million in 2022 to a whopping US$17 billion by 2032, according to data released by research firm Market.US last year.

AI is such a rapidly developing tech that production companies are constantly recalibrating their expectations when it comes to financial considerations like ROI and cost-cutting potential. (More than one producer we spoke with joked that all of their use-cases could be out of date by the time this article comes out, in fact.) Nevertheless, a picture of the impact AI can have on a production’s timeline and budget is starting to become clear.

Savvy scheduling

Applying it to roughly 20 productions, Paris-based TeamTO (Ninn, Jade Armor) has been experimenting with AI for about a decade now. Its tool uses an algorithm that can optimize a production schedule by looking at past timelines, and studio president and CEO Guillaume Hellouin estimates that it shaves about 10% to 15% in production time off the average animated series or film schedule.

There are also cost-saving opportunities related to an important but often overlooked aspect of most studio operations: the data center. TeamTO is using AI to code a program for managing the temperature of its server room to minimize fluctuations and conserve energy. This might sound like a super-technical application that wouldn’t really contribute much to the bottom line, but Hellouin predicts that the program will save the company around 15% in energy spending, and also reduce TeamTO’s carbon footprint.

TeamTO plans to share the data center program with other companies as open-source software sometime in 2024 so that the industry as a whole can move towards becoming more green. And Hellouin also sees potential in licensing its AI scheduling tools to artists on a paid model, but says this is more of a long-term plan that may play out in the next two to four years.

“During production, it’s common to have used half your budget and time, but only be a third of the way into the delivery schedule,” says Hellouin. “These tools can tell you how to optimize time spent so you can work within kids industry budgets and raise quality, without losing time and energy.”

A targeted approach

Some companies are starting to look at how AI can fill a gap in the market for production tools made specifically with animation for kids in mind. AI-powered mocap is a prime example: The licensed tools for overlaying live-action footage are typically trained off the movements of adults, not kids. And this means more cleanup work for studios.

Existing AI tools can’t handle the specific movement of animated characters, so Paris-based Animaj is building its own retargeting tool for translating live-action film into any animated character with minimal touch-ups. Without the costly retakes that traditional tools often require, mocap (and overall production) could be more scalable, says co-founder and CEO Sixte de Vauplane. Because it’s early days still, Animaj isn’t ready to say definitively how much money can be saved with AI. But the company has built several practical use-cases for its pipeline. These include text-to-3D motion (where an animator can type in a scene and create a fully editable 3D model), as well as its own “motion-to-motion retransfer tool” that uses mocap and AI to produce scenes up to 60% faster than without mocap.

This lets animators overlay animated characters onto motion-capture footage. Animaj has been developing its retransfer tool for more than six months, and it can now produce scenes that used to take 30 hours in about half that time since talent isn’t spending as much time redoing work that’s now done by technology. And it’s pairing this with AI that can read text and scan animatics to copy and reuse characters’ motions, so the team doesn’t need to film as much new mocap footage.

Based on Animaj’s early experiments and the e ciencies they’ve enabled, de Vauplane believes many studios would be well-served to earmark 10% of revenue for developing custom AI capabilities. He strongly advocates a more proactive and proprietary approach than waiting for generative AI tools made specifically for kids content to hit the market, given that this niche demand might not be a priority when every industry under the sun is clamoring for the development of AI applications. AI-powered mocap is the key to Animaj’s plan to expand existing IPs in big digital-fi rst ways. The first property it’s focusing on is Pocoyo (pictured at top), which the studio acquired in September. This preschool animation brand originated at Zinkia Entertainment in Spain and has been around for more than two decades, although it only has four television seasons to its name. (Zinkia struggled to finance new content and focused instead on repackaging existing programming.) In addition to producing a 26-ep fifth season that will be more action-driven, Animaj is working on creating several new Pocoyo specials and spinoffs, as well as short-form digital content.

A matter of policy

9 Story Media Group’s research and development team is working on developing AI tools to automate production processes like lip-syncing; animation in-betweening; rigging; increasing the resolution (a.k.a. up-rezing) of backgrounds; mocap; and predictive budgeting and scheduling. And while it hasn’t deployed any of these tools yet, the company will be posting an AI policy on its website shortly outlining for partners and stakeholders exactly how it’s using the technology.

At the heart of this statement is a vow that AI will “replace tasks, not people,” says Cathal Gaffney, Brown Bag Films’ managing director and COO of 9 Story Media Group. And the company won’t use AI to create backgrounds and character designs, either, because AI-generated images can’t be copyrighted.

The strategy behind releasing a policy that formalizes these insights is two-fold. First, it lets 9 Story inform its partners and clients (who expect privacy and originality) that AI won’t be making its shows, says Gaffney. And second, it establishes a guideline for other studios to follow, since most are also in the nascent stages of trying to figure out where and how to employ this tech. “Our policy will evolve as AI tools change, but we’ll be transparent about how we’re using it behind the scenes,” Gaffney says.

One new use for AI that 9 Story is exploring is text-to-speech for generating dialogue when actors aren’t attached and the schedule is being affected (e.g. creating animatics). AI could generate a voice for the video in minutes. In this example, tech saves time and talent costs—benefits that Gaffney believes are certain to make this use-case common-place in the industry before long.

The company’s next steps include rolling out its tools and seeing just how much AI can reduce costs—and increase profits.

More bang for your AI buck

AI technology is primed to change the industry even more than recent developments like game engines and virtual production have, says Charles Howell, president of global advertising and content at visual effects powerhouse Framestore (The Little Mermaid, Barbie, The Marvels).

Howell’s London-based shop uses neural networks to make the time-intensive process of rigging hundreds of thousands of nodes (for controlling a character’s skin and bones, for example) virtually instantaneous—a process that used to take a human several days. And Framestore has also set up groups to explore how AI can handle lighting, 2D animation and storyboarding with Previs.

Smaller studios and independent creators are also starting to play around with AI applications. For example, LA’s Encantos (Canticos) is experimenting with generative AI to build brand awareness for Skeletina, a new series it’s developing about a girl who drops into kids’ nightmares to turn them into sweet dreams. The project is based on a 2022 book from Macmillan Publishers called Skeletina and the In-Between World.

Encantos launched an AI-driven Dream Machine tool in October as part of its marketing plan for the brand. Powered by a large-language model (LLM), kids can input a written description of a nightmare into Dream Machine, which will rewrite it with a happy ending orchestrated by Skeletina. So far, 1,600 stories have been generated and emailed to kids.

Encantos co-founder and CEO Susie Jaramillo estimates that AI is saving the company around US$50,000 that would otherwise be required to hire talent to write the stories manually. This kind of spend would only be sustainable for a few months, and would be impossible to manage if there was a huge surge in submissions.

“It’s only scalable to a degree since humans are still moderating the stories,” notes Jaramillo. “But AI is getting smarter, so it can create fully original stories that are told in the character’s voice.”

Beyond this specific marketing tactic, there’s also potential for Encantos to start using AI to create new content. The studio is in the early stages of partnering with a software company to create CG-animated versions of its 2D-animated characters for the first time. Instead of having to hire a whole team, or partner with a service studio to do the expensive work of creating new character art and models, Encantos can do it with AI.

In short, the tech is enabling small companies to compete in the marketplace and promote their brands without breaking the bank, says Jaramillo.

This story originally appeared in Kidscreen‘s February/March 2024 magazine issue.