How Two Showrunners See ChatGPT Fitting Into Their Writers’ Rooms
Asked about language learning models (LLMs) at the recent 2024 AI on the Lot summit, writers and showrunners Matt Nix and Mark Groffman had a lot to say. On a panel moderated by IndieWire editor-in-chief Dana Harris-Bridson on May 16 — also featuring filmmaker Joe Penna and animator/director Momo Wang — the creatives discussed a handful of the potential uses for systems like ChatGPT in scripted projects with an emphasis on episodic work.
“Every show has instructions and rules, whether it’s a way that a character behaves or the type of world view that [the show] has or the general shape of an episode or how two characters interact with each other,” Groffman (“Limitless,” “The Umbrella Academy“) said of using LLMs to break stories. “You can feed it all of this.”
More from IndieWire
“I’ve done a lot of experimenting with my own past work,” Nix said, likening the process of teaching ChatGPT to hiring a new writer on his show “Burn Notice.” “If I just asked it to write a story for me, it was terrible. But once I gave it a very specific set of instructions and a very specific way to relate all the elements to each other, it really played into the strengths of the LLM. It was able to generate some really interesting things, which then required a lot of human interaction, a lot of editing, a lot of curating, a lot of rewriting. But in terms of what it could actually generate, it was fascinating.”
The possibility of improved efficiency in television production may or may not make a difference for studios’ bottom lines — and creatives’ job stability — as Brian Welk explained for IndieWire. Nevertheless, the near-inevitable arrival of LLMs in Hollywood portends a reimagined landscape for the craft of writing. Groffman, prone to late-night work sessions, described finding an almost idyllic partner in AI.
“It’s always awake and it’s endlessly positive and patient with me,” he said. “Those are some of the real benefits purely in the writing and creativity.”
Conversely, Nix noted that AI has the capacity to create messy misunderstandings without the human element needed to manage relationships on set. He said, “If I just trained it on the scripts, it wouldn’t know not to write scenes for those two actors who were really close when they were sleeping together and they broke up and now you can’t put them in scenes together. It would just start writing scenes for them.”
That lack of evolution, and consequential pull toward the status quo, can further impact the quality of what ends up on screen with AI naturally inclined to create more of the same with diminishing returns.
“It’s a creative process,” Groffman said. “The show that it was the first episode is not the show that it was the second episode and training an LLM it’s naturally going to spit out some sort of average.”
Plus, there are the risks inherent to AI — misinformation and bias chief among them. With the right shaping, however, Nix said LLMs are a major “evolutionary step” in the creator economy.
“Creative self-expression is built into the human nature, it’s what we want to do,” he said. “We’d like to express any of what makes us who we are and I think that self-expression has a desire to get out. … [To have good stable shows] on the air, it’s the NFL. There’s very few people who can do that well and consistently and I don’t think that’s going to change. The people who use LLMs at that level will just get that much better.”
Best of IndieWire
Sign up for Indiewire's Newsletter. For the latest news, follow us on Facebook, Twitter, and Instagram.
Solve the daily Crossword
