At the Ankler, a producer discusses what AI can and can’t do for producers. He concludes that AI can replace human editors for first-stage filtering to winnow down a huge pile of incoming scripts.
He received immediate and knowledgeable pushback from two human editors, who pointed out the fallacies:
= = = = = START QUOTE:
But as a tool for analysis, it’s a disaster because it is evaluating based on formulas, whether you’re using SAVE THE CAT or a whole bunch of screenwriting books. It’s reducing to a mean. It’s anodyne. In short, it’s got a bias against what every good story analyst is looking for: originality in storytelling and distinctiveness in the writer’s voice. No producer should be relying on AI for analysis because it’s going to miss or be unable to identify what’s unique, looking only to past patterns.
Finally, producers may only be looking to save $ on script analysis, but in doing so, they’ll put people out of work (freelance and potentially union analysts), which is evil, AND eliminate the development skills assistants and any future executives gain by doing a lot of coverage.
= = = = = END QUOTE.
Says it all. A mechanical tool looks for conformity to mechanical norms. Every measurement tool, from a scale to an AI, is designed to spot differences from a standard. This is necessary when you’re measuring an engine or a food package or computer code. It’s exactly backwards when measuring fiction.
She also hits the apprenticeship aspect, which is the worst consequence of automation that replaces HUMAN skills. Assigning low-paid assistants to do the routine parts of a task gives beginners paid work, gives them connections and familiarity with the style and jargon of the trade, and also trains them in the necessary skills. Without apprenticeships you have to rely on schools, which are utterly useless. Work is the ONLY way to learn.
Polistra's Mill 