Although a relatively new concept for many in Hollywood, artificial intelligence has been on Natasha Lyonne’s mind for years, quite possibly since she was a kid studying the Talmud at a Jewish private school in New York City.

“I feel that I understand how it formed my brain as a tween,” she says. She sees a parallel with the way religious texts are interpreted and re-interpreted and passed down: “It’s almost like one line of code as interpreted by a million different rabbis, who have theories on everything and bring the history of their entire universes, including their personal experiences, to the table, for their interpretation of a single phrase or word.”

The five-time Emmy nominee has expanded upon that metaphorical framework over the years with her own experiences, referring to an “algorithmic Tower of Babylon” that serves as her own personal creative filing system — one she says she used with her Netflix creation Russian Doll and with the upcoming film Bambo, which Lyonne will write and direct. Bambo follows a 1980s Brooklyn boxing promoter and his daughter. Lyonne is also producing through her Animal Pictures banner with Max Ferguson and Craig Mazin, Jason Weinberg and Sarah Sarando.

But the big AI project on Lyonne’s horizon is Uncanny Valley, the film she will direct from a script she co-wrote with actor-creator Brit Marling, (The OA, A Murder at the End of the World, Sound of My Voice) and scientist-author-artist Jaron Lanier. The film centers on a teenage girl who loses her bearings when a popular augmented reality game takes over her world. The title also holds a double meaning — the term is used to describe the disturbing experience of seeing AI-generated humanoid figures that don’t look quite right.

“I’ve been on this journey for quite some time, so I think I’m just a little bit more attuned,” Lyonne says of her attitude to science and AI. “I’ve read quite a lot, dropped out of Tisch for film and philosophy at 16 to become an autodidact over at the Film Forum, read and learn everything I could get my hands on.” Referring to her past experience of being in recovery for addiction issues, she says, “When I was at my lowest point, they told me to find anything, a doorknob or an ocean, and call that my power greater than me. It’s actually when I fell for science.”

While many in Hollywood and across the world are adamantly opposed to AI and the threat it poses to our environment and workforce, Lyonne has been hard at work developing a new ethical AI model through Asteria Film, the AI studio she co-founded. Adding tech entrepreneur to her resume, the actress, writer and director is among a cohort of women in Hollywood who are embracing AI, a group that has recently grown to include Sandra Bullock and Reese Witherspoon. 

I think that what people started to realize over time is that they weren’t going to be able to keep it entirely clean, so they started going entirely dirty.

Natasha Lyonne

Speaking in April at the CNBC Changemakers Summit, Bullock said of AI, “It’s here. We have to observe it. We have to understand it. We have to lean into it. We have to use it in a really constructive and creative way, make it our friend. We have to be incredibly cautious and aware of it because there are people who will use it for evil and not good. I do feel that there’s a place for it.”

Meanwhile, Witherspoon recently posted on her Instagram: “I’ve decided it’s TIME. The AI revolution has begun, and I need to learn as much as I possibly can about AI and share it with all of you. Also, FYI: the jobs women hold are three times more likely to be automated by AI, yet women are using AI at a rate 25% lower than men on average. We don’t want to be left behind. So… do you want to learn with me?”

However, Lyonne notes a harsh imbalance around the way men and women in AI are covered by the media, noting several journalists and outlets have decided to “pick on” both her and Marling, “but never on James Cameron or [Darren] Aronofsky.” Aronofsky founded an AI-focused studio, Primordial Soup, and made an AI-generated animated series, On This Day… 1776, using Google DeepMind for the visuals (the actors were human). And last year, Cameron joined Stability AI’s board of directors.

“It was really fun being attacked like that for getting involved early,” she says. “It was a little weird because, frankly, we’d been in so many meetings by that point with so many high-level think tanks around AI, because we were already sort of sci-fi futurists based on the material that was written and directed and created by us, with Amy Poehler and Leslye Headland [for Russian Doll] — this is all research we really did and had to do.”

“Everybody brings their own baggage to the table of what they think, a sort of ‘me versus the man,’” explains Lyonne of the public reaction to AI. “Everybody has a different experience of what that means to them.”

Working in part with AI research startup Moonvalley, Asteria has championed its database, Marey, as a clean and ethical AI model, which Lyonne explains is trained only on licensed data “in order to black-box my IP and be able to have output,” as opposed to the “human centipede” of other AI models that are trained on publicly available data that’s often unlicensed and unreliable.

“I think that what people started to realize over time is that they weren’t going to be able to keep it entirely clean, so they started going entirely dirty,” asserts Lyonne. “But for many of us, we were already able to read the tea leaves of the history of the cycles of showbiz. It was quite clear that things were changing. In other words, we were caught up to speed on a lot of these things, and the way it’s manifested in our business. From a broader lens, I think it’s worth noting that if we as a community do this correctly, we’re going to get to keep our movies. And if we do it wrong, well, best of luck.”

While Lyonne is embracing her own ethical AI plans, she encourages others to “keep fighting back” against the unchecked, “dirty” models of AI. And although she agrees that most forms of AI need humans to manage them, unfortunately “I don’t think it’s going to stop anyone from trying to cut jobs as much as possible.”

Last March, she was among more than 400 Hollywood professionals, including Lilly Wachowski, Alfonso Cuarón and Cate Blanchett, to sign an open letter urging the Trump administration to strengthen copyright laws, prompting the release of its National Policy Framework for Artificial Intelligence a year later, which directed Congress to provide protections from widespread copyright misuse. SAG-AFTRA has since endorsed the legislative initiative, emphasizing that members deserve protection against nonconsensual AI replications. 

However, the Trump-backed Stargate Project — an initiative intended to build infrastructure for next-generation AI — was already paving the way for a funding infusion into massive Texas data centers for AI companies like Sam Altman’s OpenAI, Larry Ellison’s Oracle and Masayoshi Son’s Softbank, which Lyonne now feels overshadowed the intended efforts of the open letter.

“We were literally, in hindsight, sort of the tinsel and the marketing department of this bigger project that was going on,” says Lyonne. “Which was really about consolidating power and healthcare and a million other things that were happening, none of which really had to do with filmmaking and copyright law.”

Right now, Lyonne says she sees the term ‘AI’ being used as “a catch-all” for anything that someone can’t understand in tech. And that lack of understanding can often be linked back to the fear and confusion around the difference between AI as we know it right now and AGI (Artificial General Intelligence)—the theoretical, potential future of AI, in which it surpasses our human intelligence and ability to reason and adapt. But she’s continually working to educate herself and understand how to use AI in an ethical manner. “I always just call and check in with my pals who are real coders,” she says. “I think there is a lot of work yet to be done.”