AI Script Coverage: Machines vs. Humans
A seasoned producer breaks down the tech's strengths and weaknesses when confronted with screening projects
Ed note: Welcome to the debut of our new newsletter, Reel AI. It seemed time we started tackling AI with a clear-eyed look at its impact on work — not through a lens of fear (valid, yes), but in a dedicated exploration of its practical applications, good and bad. After today, subscribers will receive this newsletter under its own banner (and, as always, can choose to opt out). We’re delighted to introduce Erik Barmack as our columnist (though other voices may contribute and join). Erik is a well-known 25-year veteran of entertainment and digital media — including roles at ESPN and Netflix, where he was head of international originals — before he founded his own production company. He also founded and runs a news site dedicated to AI in the entertainment industry called AI in Hollywood. Reel AI is for paid subscribers only.
I’m a scrappy guy. In 2019, when I started my own company, Wild Sheep Content, which has been focused on packaging mostly international series and films, I wanted to do it with as little overhead as possible, cutting out the red tape and focusing on creating great content with the top international talent. I had worked at Netflix for years running the international originals team, and going out on my own with fewer resources was certainly an adjustment. Artificial intelligence — yes, AI, the thing you’re likely either bullish or despairing about — has made it possible for me to stay competitive.
There’s a lot of hype about AI coming for the industry — Sora’s incredibly realistic videos putting the next generation of VFX specialists out of work, ChatGPT replacing writers, and so forth. But the reality is that most AI tools (at least, for now) are just that: Tools, not replacements.
AI is great for the really tedious tasks, like rephrasing your notes to writers so “For the tenth time, no one cares about that side character’s childhood, GET BACK TO THE PLOT” doesn’t sound so mean. You’re then freed up to work on the more creative, thought-intensive stuff. As a producer, I use AI for email drafting, image generation for pitch decks and other small things.
But as the AI-related news has been coming so quickly, a couple of things have struck me: Too much of it is either from the tech world’s point of view or it’s infused with a sense of doom. Who here in Hollywood is using all these new AI services to figure what makes sense today and what doesn’t? More important, who’s then sharing that knowledge with everyone? This is what I’ve tried to do myself for the last year, and in conjunction with The Ankler, what I now want to share with an even wider audience. (If there’s anything you’d like to see me cover, please send me an email at ReelAI@theankler.com.)
So let’s start here: Where I’m seeing the most value added from AI is in script coverage, the once time-honored starting rung in development of reading, summarizing and analyzing screenplays.
Unlike Sora creating full movies, AI coverage isn’t a hypothetical. Many large production companies are already using it, some for more than a year now. Some are integrating these script analysis tools into their development, marketing and casting decisions. These technologies are being used as a sort of co-pilot on creative decisions — and they’re only going to become more prevalent.
Speaking personally, AI coverage has really increased my company’s efficiency. We probably use it to cover about 10 scripts a week (we’re small). So at 10 scripts a week, with about three hours saved per script, over 50 weeks, that’s around 1,500 hours per year saved.
In this article you’ll learn:
Some of the more popular AI tools for script coverage (and why you may not need a dedicated AI script-reading service)
Examples of where AI coverage falls short and can’t beat what humans can do
Real comparisons of how two AI tools wrote coverage vs. the Blacklist
The types of projects best suited for AI analysis
What the future of the industry looks like with automated script coverage
Why this doesn’t eliminate the need for talented assistants
How AI script coverage can be good for screenwriters, too