AI in Action: Entertainment & Media

There’s no business like show business

This is the third in a series of articles highlighting the many applications of artificial intelligence.

Kurt Russell wasn’t always a character actor in Tarantino movies. Way back when, he started off as a fresh faced star of Disney drive-in flicks. You may remember such classics as “The Computer Wore Tennis Shoes” or “The Strongest Man in the World”. My favorite of this phase of Mr. Russell’s career is “The Barefoot Executive”. He plays a new hire at a TV network stuck in the mailroom trying to figure out how to climb the corporate ranks. His big break comes when he figures out that his neighbor’s chimpanzee has an uncanny ability to pick hit TV shows. Kurt uses the chimp to become the programming department’s new whiz kid. High jinks ensue.

If you swap Kurt Russell for a data scientist, the chimp for a predictive model, and remove the high jinks (well, most of them anyway), you have an approximation of what is going on at entertainment companies around the world right now.

Predictive Modeling

An intuitive ability to pick “hits” is a highly prized skill throughout the entertainment industry. Whether you are running a tv network, a music company, a film studio, a radio station, or a video game firm, if you can find some way to improve your “hit ratio” you’re going to greatly improve the prospects of your organization.

An example of the costs associated with finding hits is the pilot season that TV networks go through each year. Networks will pay for a single episode, a “pilot”, to be produced. These pilots typically cost between $2–5M. Out of the many pilots created, only a handful will find a place on the network’s lineup. Of these, only a few will last more than a full season. You may go through an entire pilot season, or even seasons, without finding a single hit. This is a very expensive, inefficient, process.

Improve your “hit ratio” and you’ll greatly improve your prospects.

What if instead of going through this very expensive process, you could “feed” a script and the IMDB profiles of the creative team into a neural network and get a prediction on the success of the show with an accuracy rate of 90%?

In various forms, these types of predictive models are starting to be created across the entertainment industry to try to help executives figure out what shows are going to be hits. The days of someone like Louis B. Mayer flying by the seat of their pants is coming to an end.

The need for predictive modeling is necessitated by the fact that folks like Mutt Lange or Brandon Tartikoff or Walt Disney are in very short supply. Someone who can consistently pick hit songs, or fall schedules or movie slates or video game franchises comes along a few times every generation. If you can create an effective predictive model through machine intelligence and deep learning, your success will not be predicated on finding one of these geniuses.

“Folks like Mutt Lange or Brandon Tartikoff or Walt Disney are in very short supply.”

A better predictive model not only yields better outcomes in production, it also has great value in promotion. Entertainment firms not only have to spend fortunes creating content, they also spend fortunes promoting their new content. With a better feel for what is going to be a hit, and who is going to like a certain film (demographics), you can deploy your promotional budget more effectively. This holds true in pretty much every area of media and entertainment.

Anyone who’s seen or read “Moneyball” knows that analytics has become a big part of scouting. This doesn’t apply only to baseball these days but to all the major sports. Of course, the next evolution in analytics is AI, and the thought leaders in the sports world are using AI to build predictive models not only for scouting, but for in-game decision making. Models can give you insights into tendencies, matchups, and ultimately outcomes. Sports was a latecomer to statistics, but they’ve fully embraced it now to the point where they’ve leapfrogged a lot of industries in their adoption of AI.

Natural Language Processing

We’ve previously discussed at length the value of chatbots in other fields for customer service, so let me just say that this equally applies within entertainment and media.

There are, however, unique applications of chatbots within this industry. Chatbots can be used as part of a promotional campaign for an artist, book, movie, or tv show. The possibilities are endless. A chatbot can be deployed via voice or chat, and allow a fan to converse with a character, celebrity, or group of characters. As well as a promotional vehicle, a chatbot could be monetized for established brands.

Giving a fan the chance to conduct a conversation with Sean Mendes or Spongebob or the Avengers is an experience that builds brands and strengthens the fan/character bond.

Villains like Darth Vader or Norman Bates could have entertaining chatbot “personalities.”

Through the proper curation of answers to questions, the chatbot could also capture the personalities of a character. You could have a lot of fun with classic villains like Darth Vader and Norman Bates.

This is an area that’s barely been touched, but expect to see a lot of these types of chatbots in the coming years.

AR/VR

Augmented and Virtual reality is poised to overhaul how we all consume media and entertainment. Embedded within AR and VR is a great deal of AI technology. AR/VR is dependent on core AI like machine vision and neural networks. Through AR and VR entertainment is going to become profoundly more immersive. Imagine walking through the Billie Jean video as you listen to the song, or navigating a blood vessel like one of the micronauts in the “Fantastic Voyage”. As the hardware and software improve, AR/VR is going to change the entertainment landscape to the point that our children are going to laugh at our flat screen TVs the same way we laughed at our parents’ 8 track cassettes.

“Imagine navigating a blood vessel like the Fantastic Voyage micronauts.”

A Whole New World

Sorry for the cheap Aladdin reference, but it’s pretty accurate. This article has barely scratched the surface of the many applications of AI within media and entertainment. So unless you have access to a chimp who can consistently pick winners, you may want to get up to speed on the wonderful world of AI.

AI in Action: Healthcare

AI in Action: Financial Services

Ken Tucker is a business consultant specializing in AI and Analytics. Ask him about the AI Workshop Fall Special!