Evan Agostini/Invision/AP
Picketers carry signs outside NBC in Rockefeller Center, July 17, 2023, in New York.
The basic purpose of a traditional strike is obvious: Workers, by withdrawing their labor in concert, attempt to force the business owners to pay higher wages, fund better benefits, and grant their employees some say in working conditions.
The ongoing strike against Hollywood studios from both writers and now actors is of course about pay and benefits, too. But there is another somewhat less common, and far less noticed, objective as well: The workers are trying to save the American film and television industry from the atrocious business decisions and plans of its owners.
One of the core demands of the unions, for instance, has to do with lack of “residuals” from streaming. Under the prior system of licensing shows to network and cable television, or sales of physical media, most of the production team got a cut of the proceeds, from actors to set crews. Streamers like Netflix, by contrast, refuse either to agree to similar contracts or to provide the viewing statistics that would make it possible. A recent New Yorker story details how the actors in the smash hit Orange Is the New Black made very little compared to the casts of similarly successful network shows—particularly those actors without recurring roles. Some said being on the show actually cost them money on net.
The rise of Netflix revealed a developing vertically integrated monopoly, where the studio controls both the intellectual property and the distribution network—and this time without those costly residuals. This was an enticing prospect for the other studios, but it has worked for Netflix and almost no one else. In 2022, Netflix and Hulu were the only streamers that made money—Disney+, Paramount Plus, and Max (formerly HBO Max) all lost a lot, and even Netflix lost subscribers that year for the first time.
It turns out that producing a whole platform’s worth of content costs a lot of money, as does paying for gigantic quantities of data transmission. Meanwhile, making one’s productions exclusive means forgoing lucrative licensing deals, and the multiplying expenses of so many subscriptions has driven many customers back to pirating. Keeping viewership data private, and abolishing all price signals except for streaming subscription costs, has badly eroded the industry’s ability to understand what customers actually want. As Matt Stoller writes, “In an attempt to monopolize, studio-streamers accidentally transformed a high-wage, high-profit business into a low-wage low-profit commodified one.”
This failure won’t stop the studios, of course. The end state on current trends is obvious: The streamers will consolidate into one or two huge behemoths, after which they can squeeze both their customers and their employees without mercy. Movie stars will make substantially less, a Netflix/DisneyMax+ subscription will cost $80 a month, and David Zaslav will be able to buy one of those floating aircraft carriers from The Avengers.
The core truth of filmmaking is that it’s a collaborative enterprise requiring extensive creative work at every level of production.
Now, this strike presents the studios with the juicy possibility of smashing the pesky unions forever. The goal is to “break the WGA,” one studio executive told Deadline. “The endgame is to allow things to drag on until union members start losing their apartments and losing their houses,” said another. It’s a “cruel but necessary evil,” said another insider.
Of course, this will ruin the quality of the product. The core truth of filmmaking is that it’s a collaborative enterprise requiring extensive creative work at every level of production—in writing, directing, acting, cinematography, editing, score composition, lighting, audio mixing, set design, costumes, makeup, and much more. Unions have actually been central to the success of the American film and TV industry, because they force studios to pay for the creative work necessary for a good product, and because they discipline the tendency among management (seen in every industry) to be lazy or incompetent or both.
As famed director Sidney Lumet wrote in his book Making Movies, “I’m the boss only up to a point. And to me that’s what’s so exciting. I’m in charge of a community that I need desperately and that needs me just as badly. That’s where the joy lies, in the shared experience. Anyone in that community can help me or hurt me. For this reason, it’s vital to have the best creative people in each department. People who can challenge you to work at your best, not in hostility but in a search for the truth.”
Indeed, quality problems can already be seen across the industry as labor power has declined. As author George R.R. Martin describes, streamers have blown up the previous system of training up-and-coming television and film writers in their drive to cut wages, destroying the pipeline for producing the next generation of writing talent in the process. Many films start shooting today without a completed script or without a script at all.
The visual effects industry is largely non-union, and studios commonly ask for big last-minute changes because they know artists have no choice but to work 80-hour weeks to get it done. Marvel in particular is notorious for restructuring the entire third act of its films with only weeks to go before the premiere—one reason its films have recently tended to look terrible in the final half hour. (The non-union video game industry has very similar problems of poor planning leading to brutal “crunch” periods and buggy, unfinished releases.)
Witness also how the industry has turned toward endlessly mining the same intellectual property. As Adam Mastroianni demonstrates, back in 2000 about a quarter of the top-grossing films were sequels, prequels, spinoffs, remakes, reboots, or cinematic universe expansions; that has increased steadily to about 80 percent as of 2020. (The upcoming Oppenheimer, which looks likely to become a major hit, is the exception that proves the rule.) In an increasingly concentrated industry, IP is seen as more reliably profitable because of built-in audience brand recognition and the films being similar to something that worked before—plus, the studios don’t have to pay for new rights.
What AI models can’t do at all is produce novel film or TV at a high standard of artistic quality by themselves.
This kind of lazy risk aversion has its own risks, however. It forecloses the possibility of original works that audiences might love if they got the chance. You’d never get a profoundly bizarre story like Who Framed Roger Rabbit made today, but that turned out to be a masterpiece that made back its budget seven times over. More importantly, relying on the same IP for decades also strangles the production of new IP, and eventually audiences get bored with seeing the same thing over and over again. Even Marvel is seeing declining box office receipts of late, and audiences seem sick of superhero stuff generally.
Now, studios have an answer for all these points: artificial intelligence. This has been a major sticking point in negotiations with the unions; management has flatly refused to budge on the question of where AI might be used in the production process. Executives seem to be convinced that sooner or later they’ll be able to punch in a prompt to ChatGPT (“write me a script about a cop on the edge”) and get a perfect show out the other side.
It’s true that AI models are genuinely impressive. They can conduct convincing small talk, summarize large texts with decent accuracy, produce surrealist pictures, churn out a rough draft of an essay on about any topic in seconds, lie with total confidence, and perhaps even carry out something roughly akin to reasoning. What they can’t do at all is produce novel film or TV at a high standard of artistic quality by themselves. And given how the models are trained on terabytes of human-produced internet data that is already being corrupted by AI-produced spam, there’s no reason to expect them to be able to do so in the near future. In an industry like film and TV, AIs can only be a complement to human labor, not a substitute.
At any rate, I should note that many other companies have thrived because of the high quality of their organized workforces. During the pandemic, for instance, unionized UPS did notably better than non-union FedEx because its pay, benefits, and good working conditions allowed it to hold on to scarce workers. (That of course may not stop UPS management from trying to break the Teamsters should they strike.) Some studies indicate that firms with codetermination—meaning workers have seats on the corporate board—are more productive than those exclusively controlled by shareholders.
But Hollywood is an unusual industry in how much creative work goes into it. Given the contempt for creative workers the studios now are showing, it should come as no surprise that executives are convinced a janky product like ChatGPT is already basically as good as a union writer. In reality, however, if Hollywood is to continue its international success, it is going to need lots of human workers for the foreseeable future—and to get those workers, it is going to have to pay them. Executives will have to be dragged to this success kicking and screaming.
