In 2023, generative artificial intelligence company Runway launched its AI Film Festival and received about 300 short-film submissions. A year later, the festival received over 3,000 and its 10 finalists premiered May 1 at the Orpheum Theater in Los Angeles.
In the last year, AI also has become a major point of contention in Hollywood, a topic of several labor strikes, and multiple panicked headlines. Meanwhile, filmmakers are using it.
“Among my peers, the reality is there’s more people using it than they’d like to admit,” said Paul Trillo, an AI filmmaker who spoke on a panel discussion at Wednesday’s AIFF moderated by IndieWire editor in chief Dana Harris-Bridson. “I’ve even seen people that pretend to be anti-AI, and they are using Midjourney, ChatGPT, they try and take a stance, but I know people are using things.”
Joining Trillo on the panel were Emmy-winning animator Joel Kuwahara (“The Simpsons,” “Bob’s Burgers”), tech writer, artist, and musician with indie rock band YACHT Claire L. Evans, and Runway founder Cristóbal Valenzuela.
Kuwahara said whenever he posts about AI on his Instagram, he faces a volley of vomit emojis.
“It’s almost become cool to hate AI in certain sects of the Internet,” Trillo said. “Part of me is like, ‘Oh, it’d be cool to hate AI.’”
He continued: “I always imagine AI haters smoking cigarettes and saying, ‘Fuck this.’ But it also feels like the more you hate something the more you fear something… I do encourage people that are scared to put their hands on it and experience it for themselves.”
Harris-Bridson asked the panel what they believe generative AI can do for them as filmmakers that they wouldn’t be able to do otherwise. The answers didn’t address storytelling possibilities or visual innovations but rather process. With AI, they can constantly iterate new ideas and quickly experiment with things that otherwise might have seemed too ambitious or too expensive.
“You’re learning about yourself and also being able to step outside of yourself in interesting ways,” Evans said. “I always take a discursive approach to AI, examining the data set, looking at what you can put into it, what comes out, and then seeing where you stand in relation to that output. Does it mean something? Are you projecting meaning onto it?”
Evans also said it allows her to examine her own history and find ways to “pivot and make creative course corrections” in reaction to your own past. Trillo tried to boil that idea down a different way.
“It’s given me a new process, more than anything. It doesn’t matter what the tool is. Fundamentally how I make work will be different, forever,” Trillo added, saying it’s about getting more of those “a-ha” moments that sometimes happen not frequently enough when you’re staring at a blank page. “What’s exciting is being able to fail a lot sooner.”
Kuwahara said he’s something of an introvert when he works and uses generative AI tools as a way to brew an idea as he sits in front of his computer, tinkering and improving until the model spits out the image he’s looking for.
“You can only push people so far before they start pushing back,” Kuwahara said. “Here, I can be as critical to myself and do iteration after iteration and punish myself until I can see what it’s going to be or if I’m going in the wrong direction.”
Trillo said he prefers “leaning into oddity” such as the installation “Thank You For Not Answering” or a new music video for the band Washed Out, the first ever made entirely with OpenAI’s Sora. Projects that strive to be too close to reality can be very boring, he argued; he’d rather find things that look like nothing else.
Evans, who created melodies and new sounds with generative AI tools with her band YACHT, said she’s “looking for the wonkiness,” trying to find what’s on the edges of the generative AI models and create something unusual and creative almost by accident.
That sensation was echoed in the 10 finalist short films that screened following the panel. Some were wholly generated, surreal montages of impossible images. Others, like top prize winner “Get Me Out,” combined live-action filmmaking with generative creations to tell a story about a man battling with his internal demons and grappling with agoraphobia as the actor literally wrestled with red neon monsters and the warping walls of his house.
Valenzuela said in a matter of years, AI could become an art form unto itself with interactive experiences in which people watch something be generated in real-time. There’s no language yet to describe what that might be, but it’s also not necessarily something that would compete with or replace films.
Kuwahara said he views AI as a tool that requires talented people. “I think there’s a lot of nervousness, there’s a lot of fear, and that was one of the primary reasons I wanted to get my hands on the tool,” he said. “Is this fear warranted? Is it justified? And I’m seeing first hand what it really takes to create something that is truly usable. That’s why I do think the reaction, they need to take time to play with it and experiment with it.”
