He steps off the stage at Amazon’s “AI on the Lot” and for a beat the room holds its breath. Gareth Edwards laughs, calls the tech “a billionaire second-unit director on acid,” and the applause catches in the throat. You can feel the change in the air—excitement folded with an ache of suspicion.
I watched his interview and I want you to keep one thing in mind: he’s not selling a replacement for human taste, he’s advertising a new set of creative tools. That matters because you and I both know how fast tools reshape industries—and who gets left behind.
Onstage in Los Angeles, the lights threw hard lines across Edwards’s face.
He said what every filmmaker is whispering about: generative AI is a legitimate creative partner. Edwards compared it to a second-unit director that will try anything and occasionally go off the rails. He called it “a genius at helping you,” but also admitted it “has no taste whatsoever.”
I’ll be blunt: AI is a Swiss Army knife for directors. It can speed up ideation, generate concept blocks, and produce visual references that used to take weeks with concept artists and previs teams. When Edwards suggested you should use it in development—not in the finished film—he was staking a claim that mirrors how tools usually enter studios: in the lab, quietly earning their place.
Is Gareth Edwards using AI to make movies?
He’s using it as a development tool, he says: story iterations, mood reels, variant ideas. You don’t have to accept that as the end of the line. For now, Edwards places AI in the same category as software from Adobe Firefly, image models like Midjourney and Stable Diffusion, and script assistants powered by platforms such as OpenAI. Those systems speed brainstorming; they don’t sign off the final frame.
In the edit bay, someone plays an AI test reel and the room goes quiet.
Edwards admits the machine will “go batshit crazy” sometimes, and that honesty matters. The concern most of us have is provenance: where are these ideas coming from? Are they stitched from existing films, lifted from The Creator or other indie sci-fi, or invented? That question is already reshaping conversations at Lucasfilm, Universal, and with VFX houses like ILM.
There’s a slippery side here: AI can be a carnival mirror that mangles originals—warping source material into new outputs without clear credit. Studios will demand audit trails. Directors will argue for control. You should expect legal teams at Disney, Universal, and even producers who work with Steven Spielberg to press for guardrails.
Will studios allow AI-driven ideas in big franchises?
Short answer: with conditions. IP owners will permit tool use in early stages if contracts explicitly protect source material and creative authorship. If you’re pitching a sequel to Jurassic World—a roughly $200 million (€186m) franchise—expect producers to insist on transparency about how ideas were generated and who owns them.
Backstage, producers whisper about budgets and brand risk.
Edwards floated AI as “better than CGI” for some tasks, which sounds like a headline you’ll see spun across trade sites. But you and I know that cost savings and aesthetic risk are different conversations. VFX vendors powered by Nvidia GPUs will tout speed; studios will worry about IP hygiene and long-term brand dilution.
That tension creates a practical path: use AI for iteration and discovery, then put human hands back on the wheel to craft the voice and taste. If a tool helps you find the heart of a scene faster, it earns a spot in preproduction. If it begins to do the heavy lifting without attribution or oversight, you push back.
Can AI replace directors?
No. Directors bring taste, judgment, and ethical decisions that a model can’t replicate. Edwards pointed out that AI doesn’t take notes the way humans do; it improvises. The result can be inspiring or irrelevant, and only a director can decide what stays.
I’m rooting for smart, regulated adoption. You should be, too—because this debate will determine how franchises you care about—Star Wars, Jurassic World, and beyond—are made and who profits from their ideas. Whether you trust a machine’s suggestions or not, studios will integrate generative systems the same way they adopted digital compositing: reluctantly at first, then as standard practice when the cost-benefit is undeniable.
Edwards’s stance is a provocation and a roadmap. He’s not promising a takeover; he’s promising collaboration with caveats. As someone who’s steered tentpole films and intimate sci-fi, his endorsement matters—but so do your questions about source material, credit, and taste.
So tell me: are you ready to let AI help shape the next Jurassic World or Star Wars chapter, or should studios slam the brakes until the rules are clear?