Hollywood is using metrics to decide how to do movies. Source: Study Breaks

The world of filmmaking in Hollywood is probably one area that has many what ifs. Directors and actors are always working hard to come up with film concepts and while some projects are picked up, some are abandoned or changed to suit specific standards. This intriguing uncertainty is what really makes Hollywood exciting. Fans are left guessing not knowing what’s cooking and what directors are up to. This creates the right conditions for a big hit when it comes to movies.

Despite this, the directors who invest so much money in these movies cannot live with this uncertainty. They want to have predictable outcomes before the movie is released. There’s so much money that goes into filmmaking and as such, every metric needs to be tracked in order to tell whether a film will be a hit or a flop. Such metrics may also help directors make some changes in order to achieve the desired outcome. For example, if a director feels that casting a certain actor for a lead role in a movie will make it appealing to a bigger demographic, they won’t hesitate to do it.

Filmmakers are using AI to be able to make good predictions. Source: The Verge

However, figuring out all these things is not easy. As a matter of fact, whether a movie will succeed or fail is more or less an informed guess. But things could actually change in the near future. Filmmakers are now leveraging the use of artificial intelligence to be able to make these predictions and learn where a movie will fall once it’s out. A Los Angeles based startup known as Cinelytic is developing an AI that will do exactly this.

The company licenses historical data about the performance of movies over the years and then matches the data with the themes of various films and the stars involved in them. This process is done entirely using machine learning. The data patterns are teased out and analyzed by the software and the results of potential success or failure are delivered. Think of it as playing fantasy football with a movie where you try out various stars in different roles as a simulation. The directors can swap one actor for another and get a prediction how the movie will do once it’s released.

AI will also choose which actors to cast. Source: Variety

Let’s take an example. For instance, a director in Hollywood could be working on a blockbuster movie featuring Jennifer Lawrence. The creators may not be sure if Lawrence will be a hit with fans or not so they may try out Emma Watson in a simulated role and see what the AI says. If they find that using Watson over Lawrence gives the movie a much better chance of success, they can decide to recast.

This is not the first time that startups in the US are offering this kind of software. We already have a handful of companies that are offering similar AI technology for the film industry. Hollywood studios such as 20th Century have also used AI to predict the potential success of movies after the release of the trailers. Despite this, the technology is still at the infancy stage but more could still come from it.