Natasha Lyonne is no stranger to bold moves. The Poker Face star, known for her razor-sharp wit and knack for stealing scenes, is diving headfirst into uncharted territory with her latest project: a sci-fi film called Uncanny Valley. But this isn’t just another Hollywood blockbuster in the making. Lyonne, who will direct and star in the film, is teaming up with virtual reality trailblazer Jaron Lanier and The OA co-creator Brit Marling to create something genuinely groundbreaking—a movie that leans heavily on generative AI to craft its fantastical visuals. It’s a gamble that could either redefine filmmaking or spark yet another heated debate about AI’s place in the arts.

Uncanny Valley follows a teenage girl whose life unravels after she gets sucked into a wildly popular virtual reality game. The script, co-written by Lyonne and Marling, promises a heady mix of emotional depth and mind-bending sci-fi. Lyonne will star alongside Marling, bringing her signature blend of charm and intensity to the screen. Behind the camera, she’s stepping into the director’s chair, a role she’s taken on before with episodes of Russian Doll and Poker Face. But what sets this project apart is its reliance on AI, specifically through Asteria, a new production company Lyonne co-founded with her partner, filmmaker Bryn Mooser.

In a statement that’s pure Lyonne—equal parts poetic and playful—she described the experience of working with Marling and Lanier as “endlessly inspiring.” She compared Uncanny Valley to a Wachowski sisters’ project, conjuring images of The Matrix with a twist. “Imagine if Dianne Wiest and Diane Keaton, at their loquacious best, decided to take a journey through The Matrix for sport, only to find themselves holding up an architectural blueprint,” she said.

The blueprint she’s talking about isn’t just metaphorical. Asteria is using a generative AI model called Marey, developed by Moonvalley, a startup focused on text-to-video technology. Marey is being touted as a game-changer because it’s trained exclusively on licensed material, with original creators properly compensated—a direct response to the ethical quagmires that have plagued AI development in recent years.

The timing of Uncanny Valley couldn’t be more charged. Hollywood is at a crossroads with AI, grappling with its potential to revolutionize storytelling while wrestling with its darker implications. In 2023, the threat of AI-generated digital replicas replacing actors was a key issue during the Writers Guild of America (WGA) and Screen Actors Guild (SAG-AFTRA) strikes. The unions fought hard to secure protections against studios using AI to exploit performers’ likenesses without consent or compensation. Even now, tensions remain high. Just last month, over 400 artists, including high-profile names like Billie Eilish and Mark Ruffalo, signed an open letter blasting tech giants like OpenAI and Google for claiming that training AI models on copyrighted material is a “national security” necessity.

These concerns aren’t abstract. Generative AI models, like those powering ChatGPT or DALL-E, are often trained on vast datasets scraped from the internet, including copyrighted art, music, and writing. Creators rarely see a dime, and many fear their livelihoods are at stake. The backlash has been fierce, with lawsuits piling up against companies accused of unethical data practices. Against this backdrop, Asteria’s decision to use Marey feels like a deliberate attempt to thread the needle—leveraging AI’s creative potential while sidestepping the ethical pitfalls.

Moonvalley’s pitch is compelling: Marey is a “clean” model, built from scratch with licensed footage and transparent compensation for creators. If true, it could set a new standard for how AI is used in entertainment. But the question remains: will audiences care about the tech behind the scenes, or will they just want a good movie?

Asteria, Lyonne’s brainchild with Mooser, is positioning itself as a pioneer in ethical AI filmmaking. Mooser, an Oscar-nominated producer known for documentaries like Body Team 12 and Lifeboat, brings a grounded perspective to the venture. Together, they’re betting that AI can do more than just churn out viral deepfakes or generic CGI. They want to use it to build immersive, otherworldly visuals that would be prohibitively expensive or time-consuming to create through traditional means.

Jaron Lanier, the third piece of this creative trifecta, is no stranger to pushing boundaries. Often called the “father of virtual reality,” Lanier has spent decades exploring the intersection of technology and human experience. He’s also been a vocal critic of Big Tech’s excesses, advocating for fair compensation for data creators long before it was a mainstream issue. His involvement lends Uncanny Valley a certain gravitas, signaling that this isn’t just a flashy tech stunt but a thoughtful experiment in storytelling.

Still, the road ahead is fraught. Generative AI in film is still in its infancy, and early attempts—like AI-generated shorts or music videos—have been met with mixed reactions. Some praise the technology’s ability to democratize creativity, allowing smaller studios to compete with the likes of Disney or Warner Bros. Others argue it risks homogenizing art, producing soulless content that lacks the human spark. Uncanny Valley will be a litmus test for whether AI can deliver something truly original or if it’s just a shiny new tool for cutting corners.

There’s no release date yet for Uncanny Valley, and it’s unclear whether Asteria is aiming for a theatrical debut or a streaming platform. What is clear is that the film is already generating buzz, not just for its star power but for what it represents. Hollywood’s flirtation with AI is accelerating—Disney, Netflix, and others have been quietly integrating AI into their pipelines for everything from script analysis to visual effects. But Uncanny Valley is one of the first high-profile projects to wear its AI credentials on its sleeve.

For Lyonne, the project is a chance to cement her status as a multifaceted talent. She’s already proven herself as an actor, writer, producer, and director, but helming a film that could shape the industry’s approach to AI is next-level ambition. For Marling, whose work on The OA and A Murder at the End of the World has earned her a cult following, it’s another opportunity to explore big ideas about technology and identity. And for Lanier, it’s a chance to show that AI can be a force for good in the creative world.

But the stakes are higher than individual legacies. If Uncanny Valley succeeds—both as a film and as an ethical model for AI use—it could pave the way for a new era of filmmaking, one where technology amplifies human creativity rather than replacing it. If it flops, it risks fueling the narrative that AI is a soulless gimmick, further alienating artists and audiences already skeptical of its value.

Beyond Uncanny Valley, the rise of AI in Hollywood raises broader questions about the future of storytelling. Can a machine-generated image evoke the same emotion as a hand-painted backdrop? Can an AI-crafted scene capture the messy, beautiful imperfections of human art? And perhaps most importantly, can the industry find a way to integrate AI without screwing over the very people who make movies worth watching?

Lyonne, Marling, and Lanier seem to believe it’s possible. Their collaboration is a bet on a future where technology and humanity coexist, not just in the stories we tell but in how we tell them. Whether Uncanny Valley lives up to its promise remains to be seen, but one thing’s for sure: it’s going to be one hell of a conversation starter.

As Lyonne herself might say, it’s like Dianne Wiest and Diane Keaton strutting through The Matrix—a wild, unpredictable ride that just might change the game.