Back in 2007, when YouTube and Netflix streaming was just a budding concept — and before most Americans cut their cords — the Writers Guild of America went on strike.
Among a host of demands, screenwriters were concerned that this new, mysterious way of consuming entertainment outside of live broadcast could jeopardize their job security and slash their pay.
Now, amid another Hollywood strike, screenwriters are again demanding protections around a burgeoning technology that threatens to change the landscape of Hollywood forever.
Back then, it was the residuals from content streaming on the Internet and other “new media” that topped the demands. In 2023, they're demanding that writing credit for any content be attributed to an actual human being.
“We could end up in a situation where studios are cutting writers out of the picture completely,” said Alissa Wilkinson, a film and culture critic for Vox and a member of the WGA. “And that is troubling.”
As revolutionary technologies like ChatGPT and Midjourney develop, chatter about artificial intelligence, or AI, has been everywhere lately: college students are using it to write essays; healthcare companies are testing it to help diagnose and treat illnesses; even President Joe Biden has tried it.
And for Hollywood screenwriters, who are in the midst of bargaining with the Alliance of Motion Picture and Television Producers (AMPTP) on a new three-year film TV contract, the capabilities of the technology have created a wave of panic about the fate of their jobs in the industry. It's one of many reasons writers have been on strike since the start of May.
“We're reaching the point where the goal is to cut the humans out of the process as much as possible in order to maximize profits,” Wilkinson said. “That is a dangerous place to be in.”
The WGA has proposed regulations around the use of AI that they want included in a new contract, specifically:
Earlier this month, the AMPTP issued a statement with this response:
“AI raises hard, important creative and legal questions for everyone. It’s something that requires a lot more discussion, which we’ve committed to doing.”
For some writers, the sentiment is eerily similar to the way the studios responded the last time Hollywood screenwriters went on strike in 2007.
“Last time, the line from the studios was, ‘We don’t really know how big of an effect streaming has, you don’t need residuals from that,’” Wilkinson said. “When writers hear this line from the studios they’re hearing the same thing: ‘We wanna keep talking about it, possibly because we already have plans to use it.’”
But this time around, we aren’t just talking about the distribution of media. There is a fear of screenwriters existing at all.
“The issue is, will writers exist in the future, or will it become this specialized little category that people can only afford to get into if they’re independently wealthy or don’t really need to make a living?” Wilkinson said. “I don’t really wanna live in that world.”
The potentials of AI might seem new and frightening (though it’s in every piece of technology that we use). But those who understand how it works say there is little need for writers, at least, to panic.
“AI is just math,” said Meredith Broussard, a data journalism professor at New York University. “It’s not fancy Hollywood robots.”
More specifically, AI systems create computer models based on data from various sources that already exist. That model is then used to make predictions and decisions to then generate new batches of text or images — generative media. “It’s math happening on a machine,” Broussard said.
Broussard encourages those worried about its capabilities to simply play around with the technology to understand its limits. The programs are easy to use and fun, but you’ll quickly realize that what it’s able to generate is “deeply mundane,” she said.
“If you are a business person in the studios, you may immediately think: ‘This is so great. This is free. I can replace a writer,’” she said. “Well, I'm a writer. I look at what ChatGPT generates and I think: ‘Oh my God, that is so boring.’”
If you are a business person in the studios, you may immediately think: ‘This is so great. This is free. I can replace a writer.’ Well, I'm a writer. I look at what ChatGPT generates and I think: ‘Oh my God, that is so boring.’
As far as AI replacing writing jobs, Broussard is not worried: “People working at the top of their game can produce much better stuff than any artificial intelligence can,” she said.
Still, the technology does not come without risks for the future of creative work.
Mike Ananny, a journalism professor at the University of Southern California, agrees that AI can’t replace writers: “That sort of disrespects who writers are and what writers can do.”
But he’s concerned the technology could change the creative process if studios use AI to generate outlines — or “starting points,” as he put it — bypassing the “messy” parts of the collaborative brainstorming process to maximize profits.
“Is that replacing a writer? No, but that’s where the good stuff happens — in that high risk, inefficient space of production,” Ananny said.
The result? Cheaper, more generic, less surprising content flooding our screens, he said.
“That could be a near-term outcome, where studios look at that situation and say it's relatively cheap to produce content that's ‘good enough’,” he said.
Even without AI, the streaming era’s pressure to push out more volume than ever has already transformed TV. In 2022, for example, a record 599 scripted originals hit our streaming services, according to FX — a 7% increase from 2021. A decade ago, when the streaming revolution was just taking off, FX, which has long been keeping track of such things, counted just 288 new shows.
To keep up, trends like spin-offs have become the norm. Think Wednesday, Agatha: House of Darkness and Fear the Walking Dead — shows that are sure to be consumed on account of nostalgia rather than storytelling.
“There's a whole bunch of stuff that's not great quality, but it's formulaic,” Annany said. “You know how it will turn it out.”
Annany worries that bringing AI into the fold will only make things more mundane. “If there's anything that would keep me up at night, it's that,” Annany said. “We’ll have a far simpler, more boring, less developed culture, because our culture comes from these stories.”
Beyond job security and cheapening the entertainment experience, the use of generative AI in creative works has other dangers. Because the technology uses data that already exist to create models and generate new media, it threatens to reinforce social stereotypes and biases within entertainment that many in the industry have been fighting so hard to shift in recent years.
“These systems are trained on data scraped from the web. The web is equal parts wonderful and highly toxic,” said Broussard, who recently published the book "More Than a Glitch: Confronting Race, Gender and Racial Bias in Tech."
In April, an investigation by the Washington Post revealed that a popular dataset used to train AI — the Colossal Clean Crawled Corpus, or C4, which is assembled from more than 15 million websites — used a white nationalist site, VDARE, in its database, as well as other disreputable sites.
“Every kind of human bias you can imagine is in the output of these generative AI systems because it's in the inputs,” Broussard said.
No matter the risks, generative AI has triggered a seismic shift in Hollywood, and is here to stay.
In March, the fourth version of ChatGPT was released — ChatGPT 4 — which, unlike its predecessor ChatGPT 3.5, can take images as well as text input.
While the outcome of the WGA’s contract remains unknown, this paradigm-shifting force is sure to change the landscape of Hollywood — and almost every other industry — forever.
Moving forward, Annany said, we must think critically about how to coexist with it in Hollywood and beyond.
“Know who you are and what kind of creator you want to be,” Annany said. “Ask yourself: Is this technology bringing you closer, or further away from the world you want to live in?”
