The biggest advantage of the technology, on its face, seems to be its ability to predict how well the project may do at a box office, based on legacy data. The biggest, more important disadvantage remains its inability to predict a project’s perceived value over time.
Of course, the marketability of a script remains important. But the way a production is marketed can influence box office results despite a script’s metrics (e.g., number, length, and/or gender type of dialogues; screen time per character, etc.).
I’d prefer an AI system that helps writers conceive the script in the first place. For example, a great help would be script analysis that pings an already produced idea or scene–or even one that’s currently in production. With the democratization of filmmaking, who can (or would dare) claim to have consumed every piece of content for a particular genre (i.e., Romance)?
What’d also be nice is an AI product that helps writers develop a script. What combination of ideas do audiences want to see, say, based on search engine analysis? What kind of events should a writer consider or eschew?
Now I know what you’re thinking: “Just who’s writing this thing anyway–me or it?” The time may have long since come to rethink what it means to “write” a screenplay…much the way producers have had to rethink what it means to produce a project; actors, what it means to portray a character via computer-generated imagery (CGI).