A New Tool in the Writer’s Room
A ‘story pass’ is a term used by screenwriters for a specific editing phase where the script is reviewed with a single focus, such as character, theme, or plot. A ChatGPT story pass, or more broadly an ‘LLM story pass,’ adapts this concept for the age of artificial intelligence. It’s the process of feeding a screenplay, treatment, or even a rough idea into a large language model (LLM) like ChatGPT, Claude, or Llama to receive analytical feedback and creative suggestions. This is not about the AI writing the script from scratch; rather, it’s positioned as an tireless, data-driven brainstorming partner that can help a human writer identify weaknesses and explore new possibilities.
The writer might prompt the model with specific requests, such as: “Read the following script and identify any potential plot holes or continuity errors,” “Suggest three alternative third-act turning points for the protagonist,” or “Does the antagonist’s motivation feel compelling? If not, why?” The AI can process the entire text contextually and provide feedback in seconds, offering a new and often unexpected perspective on the material.
Potential and Pitfalls
The promise of this technology is its ability to augment the creative process. For a writer struggling with writer’s block, an LLM can generate a dozen loglines or character concepts to get the ideas flowing. For a nearly-finished draft, it can serve as a tireless proofreader, checking for pacing issues or inconsistent character voices. The key advantage is speed and volume—an AI can simulate a small writer’s room, offering multiple angles on a creative problem almost instantly.
However, the pitfalls are significant and subject to intense debate. The primary concern is the risk of producing generic, formulaic work. LLMs are trained on vast datasets of existing stories, and their outputs are, by nature, a sophisticated remix of what they have already seen. This can lead to suggestions that feel predictable or derivative, reinforcing existing tropes rather than creating something truly original. Furthermore, these models lack genuine human experience, consciousness, and emotion, which are the bedrock of compelling storytelling. An AI can identify that a scene is structurally intended to be sad, but it cannot feel sadness or understand the subtle, idiosyncratic nuances that make a moment truly resonate with an audience.
The Industry’s Skeptical Embrace
The advent of AI in screenwriting became a central issue in the 2023 Writers Guild of America (WGA) strike. The union fought for and won protections that regulate the use of AI. The current agreement stipulates that AI cannot be used to write or rewrite literary material and cannot be considered ‘source material,’ thereby protecting writers’ credits and authorship. A writer can choose to use an AI as a tool (if the company allows it), but a company cannot force a writer to use an AI. This reflects a deep-seated fear that studios will see AI as a way to cut costs, devalue the role of the human writer, and create a future where scripts are assembled rather than written. While the technology is a powerful new tool, its place in the creative hierarchy of Hollywood remains one of the most contentious issues facing the industry.