We can't talk about AI

The last few years in the VFX industry have been a real roller coaster:

• During the pandemic, demand for artists surged, coinciding with the launch of several streaming platforms and a volume of content production like nothing we’d seen before.
• The writers’ and actors’ strikes in Hollywood halted a huge number of productions, and at the same time the big studios started cutting costs: they opened sites in India, closed in “expensive” countries or those with few tax incentives, like Singapore, and reduced outsourcing to smaller studios.
• Right now we’re in recovery: productions seem to be coming back, especially in London. The big companies are getting overwhelmed again, passing some of the work to smaller studios and even hiring remote artists. There’s a sense of normality again.

But a rumor, a shadow, looms over the industry.
And we can’t talk about it.
At least, I feel like I can’t.

It’s generative AI. Many people outside the industry probably think visual effects and film have jumped on this new wave and that we’re churning out content like crazy. The truth is very different.

There’s a mix of rejection and denial among artists. It’s a genuine taboo. Most of my colleagues don’t want to talk about it, don’t want to learn (many would rather do something else entirely than generate content with AI) or simply aren’t aware of how fast everything is moving.

It’s true that most studios are starting to use models to generate “non-creative” utility passes: depth channels, normals, rotoscoping or paint. Work that is hardly creative and would normally have been very tedious to do by hand.

But the area that is advancing the most, the one that terrifies and fascinates me in equal measure, is generative AI. I know the Internet is filling up with AI slop, that deepfakes and fake news are dangerous, and that most models haven’t been trained ethically. I think the only thing saving the industry from being swallowed up by AI is the massive rejection from the general public.

But you can’t put the genie back in the bottle. It’s only a matter of time before fully generated content eats an ever bigger slice of the pie. It will start with advertising and music videos, then move on to TV series and low-budget or independent films, and finally it will reach cinema.

And I’m sorry if this sounds pessimistic, but I need to talk about it.

Lately, every time I go to dailies, the daily review sessions, I savor the chance to talk with the supervisors about the creative aspects of the shots. Sometimes I sit in silence and appreciate how much I love the industry and the creative process, and I wonder how long it will last and whether I’ll be able to adapt to the change that’s coming.

The film industry is suffering a genuine creative crisis: it’s unable to take risks on new stories. I miss the cinema of the 80s, more creative and full of real adventure. Sometimes I think generative AI might give wings to new projects that otherwise wouldn’t see the light of day, serving as a kind of beta test before bringing them to the big screen the traditional way.

I believe there is real creative talent using these tools. One example is the Spanish studio Contanimation. I love watching their work, even though I’m aware that supporting and enjoying it is frowned upon. And that perhaps, someday, my industry will disappear as we know it today.

I feel the big companies are very slow to adapt, that we artists aren’t dedicating the time we need to retrain, and that not being able to talk about it only makes things worse.

I’d love to hear your point of view:

• Do you feel generative AI is a taboo subject?
• How do you think it will affect the VFX industry?
• Are you afraid, or excited, about what’s coming?