Introduction
With technology and storytelling deeply interconnected, the entertainment industry finds itself in a period of transformation triggered by artificial intelligence. Rachel Joy Victor, co-founder of FBRC.ai and adjunct professor at the University of Southern California’s School of Cinematic Arts, is helping shape this frontier. From co-producing events like AI on the Lot to mentoring students and startups alike, Victor offers a well-informed perspective on how AI is reshaping the film industry.
“AI was always on my radar,” Victor recalls. “I studied computational neuroscience as an undergrad, so I was already working with AI and robotics well before generative tools were usable. I incorporated machine learning into various projects, especially during my time designing immersive and interactive experiences at a brand innovation studio.”
“But the generative tools weren’t quite there yet,” she notes. “When generative AI improved, it finally felt like we could design systems powerful enough to be transformative. That’s when we started FBRC.ai and dove fully into this space.”
Educational Observations
In discussing education, Victor observes how massive changes are reshaping both the film industry and the way it is taught, which go far beyond AI.
“Film schools are still figuring it out,” Victor observes. “It’s a challenge because a lot of things are converging at once. In my class at USC, we discuss how the economic foundations of content and its distribution have undergone major changes over the last 10 to 15 years. That’s disrupted the industry in ways we haven’t fully addressed. A lot of production models just aren’t tenable anymore.”
“People focus on things like LED walls for virtual production, but the real change is that production is becoming more data-oriented,” Victor adds. “What’s happening with AI dovetails with that.”
People focus on things like LED walls for virtual production, but the real change is that production is becoming more data-oriented.
“I wrote our FBRC.ai industry report that we released two months ago, and there’s a section where I describe filmmaking as ‘I capture footage and stitch that together,’” she says. “Then at a certain point, it evolved to ‘I’m not just capturing video, I’m capturing data, camera position, motion, etc.’ All of those elements get reinterpreted during a complex production process, and AI is another layer of processing.”
AI Tool Recommendations
As part of her role and driven by her curiosity about technology, Victor has had numerous opportunities to review AI tools. “Most of the off-the-shelf tools are not necessarily the way to go because they’re trying to be accessible to everyone, but at a cost of finer control,” she observes.
“Model-wise, Veo 2 and 3 have impressive visual results, and are considered clean models in terms of copyright,” Victor adds. “Kling also comes up a lot for its lip-syncing capabilities. ComfyUI is where everyone is congregating because you can use multiple models and different control methods.”
“A lot of individual tools that support content creation workflow are also coming out, like ElevenLabs as the standard for voice generation and TwelveLabs for quickly searching your content,” Victor says. “Across the spectrum of tools, there are exciting things that are happening, but there’s nothing that’s established itself as the singular tool you must learn. There are principles and best practices of prompting and using ComfyUI that should transfer well to whatever else comes out next.”
Opportunities in AI
Major technologies, such as AI, bring both disruption and opportunity. “I don’t think the prompt engineer role is where the most value is,” Victor says. “To see real AI adoption, three things need to improve: the models themselves, the tools that implement them, and education: how people learn and integrate these skills.
“People should incorporate AI competencies in their broader skill set. Existing skills, like VFX expertise, remain valuable; AI just adds to them, and there’s still a lot of open ground. ComfyUI, for example, is widely used but not yet stable for scaled production. For studios, there’s still a lack of quantification and a lot of opportunity.”
For studios, there’s still a lack of quantification and a lot of opportunity.
On the individual level, Victor believes it’s essential to broaden skill sets. “The probability of landing a job coming right out of film school, especially as the amount of original development that’s happening within Hollywood continues to scale down, is lower.”
“There’s value in people who can create good images and have art direction sensibility,” Victor says. “At the studio level, they’re looking for people good at building ComfyUI workflows, or go a level deeper to Python programming, or build integrations with other tools.”
Future for Studios
Movie studios are currently watching and experimenting with AI to understand how and where their businesses will evolve competitively over the next few years. “I think the hardware is always going to be important, especially at the larger studio level,” says Victor. “It makes sense for them to want to own a lot of things locally, both for security and safety.”
“There is still value in the cloud side of this, especially for remote teams,” she continues. “There are a lot of emerging smaller teams that want flexibility and the ability to collaborate even if they’re in different locations.”
Intellectual property is inexorably connected to the hardware question, as Victor notes, “The Lionsgate–Runway deal is interesting, and we may see more examples like that. Studios with large catalogs, if they can determine the rights, which often have complex sub-ownership, will likely want to train their own larger models or LoRAs based on specific characters or styles. They’ll want to do that locally, both for control and for safety reasons. But then they may choose to make those models securely accessible to their distributed teams for productions, via cloud-enabled workflows.”
The Next Steps
Asked to consider what might happen to the movie business in the coming years with the advent of AI, Victor has some observations. “An individual can do a lot with these tools, but there’s still value in teams,” she says. “The best productions usually involve multiple people because of the variety of skills and the scale required. Asteria has been transparent about their process, for example, with Cuco—and that helps show that collaboration still matters.”
“Studios are still important. They act as banks and IP owners and have established distribution and marketing pipelines,” she adds. “Those roles will be reimagined, but they’ll remain valuable. We also need to keep interrogating the distribution model. There’s experimentation happening—what goes to theaters, what streams, etc.”
“You see it with Sinners getting a 70mm re-release, or Twisters in 4DX. The role of theaters is evolving because the experience is different,” says Victor. “Tools may speed up production—an individual or small team creates something, and within six to nine months, a polished version gets released theatrically, then continues digitally. That back-and-forth is growing. The term ‘transmedia’ comes and goes, but the concept—building franchises across various formats—is becoming increasingly relevant. That’s something that will evolve a lot in the next five years.”
The term “transmedia” comes and goes, but the concept—building franchises across various formats—is becoming increasingly relevant.
To learn more about Rachel Joy Victor, please visit her website at: https://www.racheljoyvictor.com and https://www.fbrc.ai/
