Here’s How AI Is [Eventually] Coming For Your Animation Job
If you work in the animation industry, how worried should you be about AI stealing your job? Well, the answer isn’t a straight one unfortunately, but that doesn’t mean that AI isn’t coming for your job; it just means it isn’t coming for it in the way you may think.
Artificial Intelligence or ‘AI’ is a trendy buzzword to describe a skill that computers have had for decades. (If you’ve ever used a filter in Photoshop, then you’ve used ‘AI’ on a more basic level.) The only thing that’s changed is the massive computing power that’s become possible in recent years, and the ability to parse enormous volumes of data that such power provides.
Years ago, in an article I can no longer find, the author made a point of discussing Michael Bay’s approach to CGI in his Transformers films; in other words, the excessive amount of it. The one point that struck with me from that article was the one that said something to the effect of ‘if you can be replaced by a computer, you will be replaced by a computer’. A scary thought to be sure, and it’s true for every job besides the ones that are physically demanding (although technology has come for those in the past too).
We’ve seen it before
So that got me thinking about how a computer could come for workers in the animation industry. But first, a caveat: we’ve been here before! The prospect of technology cleaving off whole departments is nothing new, even within animation. The Xerox machine (photocopier) took over the role of the inker in the 1960s. Two decades later and it was the painter’s turn as CAPS was developed. Since then there’s been the transition from traditional hand-drawn animation to 2-D and 3-D CGI, and digital projection has supplanted 35mm film.
What’s important to remember is that the Xerox machine, CAPS, and CGI software such as Renderman, Adobe Flash, etc. are all just tools. While they did replace individual workers in roles, they did not replace all of them, and did not do so all at once; Disney animators working on TRON would not become “obsolete” for more than a decade and even then hand-drawn animation continues to be made. These tools also opened up new roles that did not exist before. (Here’s a thought: does a second of hand-drawn animation require more or fewer people to create than one second of CGI animation?)
Is AI simply a tool or is it something else though?
One the one hand it is definitely a beneficial tool to some. Need a story treatment but stuck with creative block? ChatGPT could probably throw you a lifeline with an idea. On the other hand, it is definitely a threatening tool to others. Visual development artists will find intrusions into their work if they have not done so already. After all, why pay someone to create a visual concept for a show from scratch when you can just throw your desires into a machine and have it do the hard work for free? And if you’re a writer, what can help you in one way could hinder you in another as studios no longer look for you to write scripts so much as finesse what a machine drafted first. UPDATE: as of writing, the WGA has decided to strike with one of their concerns being the use of AI to replace writers, and the use of writer’s creations in the training of AI models.
My point is that the ‘AI’ is going to be both a benefit and a threat. The low hanging fruit of static art, music, and the written text is being picked at now. Motion pictures will come later this decade.
How worried should you be?
Perhaps not as much as the media and AI hype men are making out you should be, but also not enough that you can ignore it completely. A computer can apparently write text in the style of a human sort of well, and can even create a new song in the style of Drake. But the text isn’t perfect, and the fact that AI can create a song by a top artist probably says more about the sorry state of popular music than it does about its creative chops.
From the consumer’s perspective it’s a different story. The possibilities include your favourite song being used to extrapolate and create additional, AI-generated variations. As I wrote back in 2016, a company like YouTube is ideally situated to exploit the audience’s demand for more of the same based on the amount of data they collect. It is not a far stretch to see episodes or shows tailor-made for you and you alone. AI will get us there and satiate the intense desire for entertainment.
Except that there is a difference between the kind of content that YouTube contains and more traditional channels. YouTube is the fast food to Disney’s four course meal. As I noted in 2016, a hefty chunk of the audience is more than happy to subsist on fast food with the occasional more substantive meal.
Nibbling Around the Edges
Ultimately what we are going to see is AI nibbling around the edges. Scripts will get some help, visual development will get a shot of interpretation, animation itself will get some AI-infused movement and/or automation (i.e. no more inbetweeners), even musical scores will go through the AI ringer. It isn’t all going to change at once; especially since AI art has a way to go before it can even accurately recreate something that already exists.
Think of it as akin to the switch from hand-drawn to digital animation. A whole slew of young artists got a headstart by only learning digital. Many older workers were able to transition with a bump or two. A few masters couldn’t make the leap but while their skills were no longer useful to production, their knowledge of technique was immensely valuable to younger artists. AI technology in production will follow a similar pattern.
Preparing for the AI Threat
How can you insulate yourself against the AI threat? On the one hand, you can’t, unfortunately; job opportunities will just get fewer and fewer. What you can do, though, is learn whatever new skills are coming to the fore. Check job listings to see what skills employers are starting to look for. E.g. AI can spit out stories, but whoever can input the best prompts and get the best results will get a gig. Knowing how to pull the right levers in the right way in Stable Diffusion will give you an advantage over someone who does not. Look at ways your current skills can translate into new ones. Seek out training or certification on new technologies in the same way you do for animation software now.
A saying I’ve come to respect is the one that says ‘technology never goes backward’. That is, once a technological advancement is introduced, there is no going back. The earth may be dying, but nobody is pulling the air conditioner out of their house. Once people saw talkies, they weren’t going to content with silent films any longer. The same is true with AI and animation: as it changes the production landscape, nobody will go back to relying on manual processes. They either won’t commit the time and resources, or they simply won’t be satisfied with the results. That’s disappointing to write because the immense human creativity out there will take a hit, even if it is ultimately allowed to shine through. It is, however, the best way to understand how change will affect you.
Fifteen years ago, the WGA went on strike partly because they perceived the rise of streaming would have an adverse effect on their member’s incomes (writers at the time did not receive residuals on streamed content; only broadcasts and home media sales. This time around they (personal opinion) are attempting to close the door after the horse has bolted with regard to AI use. A resolution will be reached or course, but will the AI can simply be kicked down the road (for studios to exploit in the interim) or will we see writers having to embrace it?
It’s hard to say exactly when AI will consume current jobs and roles, but preparing for the inevitable is the best course of action. How will you do so?
As a parting message, here is a film created to document the last day of using hot metal type to create The New York Times before the newspaper switched to digital typesetting. It’s a fascinating documentary on technological change: