On 2 May 2023, it was pens (or laptops) down for 11500 WGA members following an overwhelming strike authorization vote. Negotiations abruptly ended when AMPTP (the industry body representing the main studios) failed to counter an offer by WGA. Writers and their allies (including those from other entertainment unions such as IATSE and SAG-AFTRA) vociferously took to the picket lines and social media with witty placards and chants proclaiming they will not back down as they face an existential crisis in what was once a comfortable career in favor of increasingly poorly-paid gigs where professional writers could no longer support their families amid skyrocketing studio profits.
Similarly, the main actors’ union SAG-AFTRA asked its members to vote on strike authorization ahead of its commencement of contract negotiations with AMPTP on 7 June. They have since secured that strike vote as a bargaining tool. The DGA, which represents directors, opened contract negotiation talks on 10 May with AMPTP, but its official support of the current WGA strike is unclear.
WGA and AMPTP haven’t reconvened since the strike began, so the issues are far from resolved. It’s likely, they won’t do so until after DGA and SAG-AFTRA negotiations are complete. The strike can legally last for three months, after which AMPTP can declare a force majeur and cancel all existing labor agreements with WGA. The previous strike of 2008 – 2009 lasted a hundred days and the current standoff appears likely to follow a similar trajectory.
The Role Of AI
The WGA negotiating team has highlighted two key areas of concern – the (ab)use of AI and TV writers’ mini-rooms. The former is a concern to all writers especially in terms of valuing their craft and defining authorship of a film or TV series.
The term “creator of a literary work” has a specific legal meaning. A creator must be human and the work must be in a tangible fixed form such as a screenplay. AI is still in its infancy, so its ensuing capabilities are yet to be witnessed. WGA insists some ground rules on these technologies must be set during the current contract negotiations to protect its members from abuse. Otherwise, writers may face the same problems as cryptocurrency and social media users because these platforms weren’t subjected to early regulation. Cryptocurrency was never intended to allow the criminal black market to flourish and social media was certainly not created for the spread of disinformation and hate speech.
Writers also argue that AI and other large language models do not actually create original stories. They simply gather big data from a multitude of sources, then process, interpret, and repackage it. It’s a form of plagiarism.
Studios decry that writers are Luddites and will go the way of the horse and cart if they don’t embrace new technologies and modern business practices.
Writers are not inherently anti-progress or hate technology. AI, algorithms, and big data certainly have a role to play in the telling of stories. Much like a driverless car, AI gathers and processes terabytes of data to give the car instructions on how to drive. However, a human needs to drive the car in the first place in order for it to gather the relevant data.
AI learning is based on binary math. A red light means stop, a green light go, a speed sign means the car needs to check its speed and adjust if necessary. The more it drives, the more “human-like” its decisions become through “learning.”
In a possible writing scenario, a studio can ask AI to write a funny sitcom without actually hiring a writer – and it probably can, based on painstaking computer analysis of popular comedy shows, ratings, and the jokes it contains. But is it any good?
AI can only determine if a joke was funny or not based on the data it’s fed. There isn’t any inherent cultural context or social nuance driven by a human in AI-generated scripts. It’s the difference between speaking to a live customer service operator and scrolling through a maddening menu of pre-recorded options. You might be allowed to speak to a live human after all other options have been exhausted.
The issue of plagiarism can also be offset by allowing random variables to enter the AI mix. Imagine if AI was asked to write a breakfast table conversation of a Mid-Western family with traditional values. Al can conceivably insert random dialogue from a magazine article to avoid the human writer interface. It may also swap out words with synonyms to avoid detection by the plagiarism police.
An AI-generated sitcom will be more precise if the user specifies a workplace sitcom with a young, hot, urban, ‘nepo’ baby cast that work in an influencer advertising agency, than simply keying in “write a sitcom.” But it will never be more than a simulated human interaction, no matter how advanced the technology becomes. If we reach the stage of “I can’t believe it wasn’t written by a human,” where are we as a species?
Actors are facing similar concerns with talent agencies increasingly requiring actors to submit to a face scan and digitized voice recording to expedite reshoots, promotion, or spinoffs without further compensation.
Mini-Rooms
This is the other major contentious issue that WGA is fighting to address. Mini-rooms are a means for the senior writers of a potential TV series to convene and present an extended pitch to a studio to give them a better visualization of a potential film or TV show.
Imagine if a local municipality requested tender documents to build a school or hospital. Typically, these require developers to submit plans, models, budgets, timelines and other pertinent information. What if if they were told to lay the foundations and erect the building frames to demonstrate they are capable of the job before being paid, without any guarantee of winning the tender? This is the crisis writers are facing in unregulated mini-rooms.
Writers (generally senior) in mini-rooms are often paid at WGA scale without any guarantee that the show will be greenlit or that they will even be staffed. Mini-rooms have increasingly become sweat shops where writers are expected to front load a TV series as much as possible. Sometimes, they are even asked to provide a pilot episode script or a comprehensive season outline in a matter of weeks so studios can assess a project’s viability while saving the studio money on development costs.
If a show is greenlit, studios argue that fewer writers are subsequently required in the writers’ room since a substantial amount of work has already been completed in the mini-room. Again, this is another cost-cutting measure at the writers’ expense. WGA is insisting on a minimum number of writers per TV show – one writer per episode for up to six episodes and two writers per episode after that until twelve episodes.
AMPTP is vehemently opposed to this archaic minimum staffing formula as being incompatible with the flexible and changing nature of creative industries. They call it ‘featherbedding’ (an old union practice allowing members to get paid for doing nothing). This practice can conceivably reduce the minimum number of writers to zero per script with the help of AI, or one writer per multiple scripts to do a polish or rewrite as a work for hire rather than as a creator. They also posit that showrunners should decide how many writers they need, especially in the case of auteur-driven TV shows which are often created and written by one writer.
Television seasons are getting shorter – down to six or eight episodes from twenty-four per season and writers are increasingly being pushed into day worker rather than staff status. Studios also like the practice of contractually taking writers off the market in case rewrites or polishes are required, or a subsequent season is ordered. This is fine if they paid writers to be off the market, but they don’t.
WGA wants to set some guard rails and basic protections for their members to ensure that writing is a viable and sustainable career choice. Time will tell the final outcome of this protracted labor dispute.