Hollywood Professionals Are Quietly Training the AI Models That Could Replace Them

Hollywood Workers Are Moonlighting in AI Training to Survive

When thousands of writers and actors walked picket lines in 2023, one of their loudest fears was that generative AI would hollow out the entertainment industry. Now, three years later, that fear hasn't faded — but some of those same industry workers are quietly helping to build the very technology they once protested, simply because they need the money.

As Hollywood's job market continues to shrink, a growing number of film and television professionals have started taking on AI training work on the side. A handful have even gone public with their experiences, shining a light on a practice that remains deeply controversial within the creative community.

Writers, Editors, and Executives Turning to AI Gigs

In May, screenwriter Ruth Fowler — known for Little Disasters and Rules of the Game — published a candid personal essay in Wired detailing her experience taking AI training jobs after entertainment work dried up. She was straightforward about her motivation: she needed money for rent and groceries. That same month, TV movie writer Robin Palmer, whose credits include work for Disney Channel and Hallmark, spoke publicly with CBS News about her own involvement in AI training — acknowledging that some of her peers might view it as crossing a picket line.

Editor Gabe Sena is another entertainment professional who has taken up AI training as a side gig. A UCLA film school graduate who typically edits documentaries and videos for nonprofits such as the Make-A-Wish Foundation and the Children's Defense Fund, Sena says curiosity — not desperation — initially drove him to explore the work during a slow summer in 2025.

"I'm mid-career and I don't want to become irrelevant in my field," Sena told The Hollywood Reporter. "This technology seems like a black box to a lot of people outside the tech industry, and it made more sense to me to immerse myself in it rather than simply complain that it exists."

Sena found his way into AI training through Mercor, a recruiting platform founded in 2023 by three college dropouts who are also Thiel Fellows — participants in billionaire Peter Thiel's program that funds and mentors people who leave or skip college to pursue entrepreneurial ventures. Backed by major venture capital firms including Benchmark, General Catalyst, and Menlo Ventures, Mercor describes its mission as "organizing human intelligence to power the AI economy." The company raised $350 million in a Series C funding round in 2025, putting its valuation at $10 billion.

Bound by a non-disclosure agreement, Sena can't reveal the specifics of his assignments, but he described one project that involved evaluating how well a generative AI system's outputs matched the prompts it was given. He's cautiously optimistic about his own future, believing that adaptable and skilled workers won't be fully replaced — even if some tasks get automated away.

"There are parts of my job I'd gladly hand off to a smarter system," he said. "But even the most routine-looking tasks in my work involve a lot of judgment and subjectivity."

From HBO Executive to AI Trainer

Steven Woolworth, a former development executive at HBO and Studio TF1 America, took a similar leap after spending an exhausting year and a half searching for traditional Hollywood work without success. A friend who is a Writers Guild of America member forwarded him an email from Mercor, and Woolworth decided to take the plunge.

Despite being a vocal supporter of AI regulations and safeguards in the entertainment industry, Woolworth framed his decision as a pragmatic one.

"I could keep my head buried in the sand, or I could step inside this world and gain a real, firsthand perspective on how AI training actually works — and perhaps help shape it in a way that's healthier and more positive," he explained.

The work, he said, has "kept a roof over my head for the past year," and he remains unconvinced it could fully replace what he used to do. Discovering new talent at film festivals, attending live theater, building real human relationships — these are things AI simply cannot replicate, he argues. "I know today it can't do the essential human things that I do," he said.

How AI Training Actually Works

The technical process these workers are participating in is called Reinforcement Learning from Human Feedback, or RLHF. It works in three stages. First, human reviewers score an AI model's outputs — for example, judging whether a fictional character's reaction to bad news feels authentic. Those scores build a training data set. The AI then uses that data to develop a second "reward model," which is in turn used to retrain the original AI — gradually removing the need for direct human involvement.

For the people doing this work day in and day out, it can be a mind-numbing experience. One veteran writer compared it to sitting in a high-school exam room with a strict proctor, watching an endless parade of AI-generated responses flash across a screen and rating each one. What initially felt like guiding a curious, well-meaning machine quickly turned into something far more tedious and psychologically draining.

"How many times can you tell a machine it's wrong without losing your mind?" the writer recalled.

A Growing Market — With Real Risks

AI training is increasingly positioning itself as a field that values specialized expertise. Companies are now actively recruiting professionals — not just in entertainment, but also in medicine and law — who can earn upward of $100 per hour helping AI systems sound more human and correcting domain-specific errors. According to data from the job platform Indeed, AI-related job postings in the arts category doubled between May 2025 and April 2026, climbing from roughly five percent to nearly eleven percent of total listings. As of publication, platforms like Handshake were advertising rates of up to $44 per hour for creative writers and up to $100 per hour for music professionals with advanced degrees.

However, working conditions in this industry are far from ideal. Workers have reported unstable employment, cutthroat competition for assignments, rushed timelines, and little transparency about who the end client actually is. Two major RLHF providers, Surge AI and Scale AI, have each faced legal action — the former accused of unpaid wages and worker misclassification, the latter of exposing workers to deeply disturbing content.

The Moral Dilemma Dividing Hollywood

Within the entertainment community, opinions on this phenomenon are deeply divided — though even critics tend to express empathy for those taking the work.

Breaking Bad creator Vince Gilligan offered a measured response when asked about it. "I've been extraordinarily lucky throughout my career, so I'm not going to judge folks who are simply trying to provide for their families," he wrote in an email to The Hollywood Reporter. "I just thank god I've never been faced with that choice myself. And I appreciate that my career started in 1990, not 2026."

Storyboard artist Sam Tung, whose credits include Citadel and Twisters, echoed that sentiment. "The people taking this work aren't doing it because they want to undercut other human workers. People are really struggling right now, and they have mortgages to pay and kids to feed."

But Tim Friedlander, a voice actor and president of the National Association of Voice Actors (NAVA), urges caution. While AI training opportunities are multiplying on platforms commonly used by voice talent, a recent NAVA survey found that approximately 20 percent of respondents have already knowingly lost jobs to AI tools. For Friedlander, the short-term financial relief these gigs offer may come with consequences that are far more painful in the long run.

The tension at the heart of this story is one with no easy resolution: workers caught between immediate financial need and the unsettling possibility that the work they're doing today may contribute to making their own skills obsolete tomorrow.