By Jason Friedlander, Director of Product Marketing
Three people debate a tense love triangle in a futuristic office. Suddenly, a hole to outer space opens up in the floor. A man who has just spat up an eyeball steps through.
Sunspring, a science fiction short film based on this premise, could very well be an experimental film written and directed by a group of avant garde cinema school students. Indeed, the film was judged to have creative merit, placing in the top 10 of Sci-Fi London’s 48-Hour Film Challenge, a contest in which participants create a film based on prompts over the course of two days.
There’s just one catch: Sunspring was written by artificial intelligence.
The robots are coming and, in the not-so-distant future, they’re going to change the way we make, program and watch TV and film. While we’re growing more used to the idea of AI completing practical or empirical tasks, common sense says the world of human creativity should be beyond the grasp of artificial intelligence. But AI is already expanding into the TV and film industries, and we’re bound to see a rapid rise in the use of AI technology in the next few years.
For many creative professionals in TV and film, this means one thing: major industrial disruption.
Already, the alarm bells are ringing. On Reddit, forums of assistant editors worry they’ll soon lose their jobs. Meanwhile, over at Screencraft, a screenwriting craft and business website, a scathing review of the AI writing in Sunspring reveals an underlying anxiety that human screenwriters will soon be obsolete. As film editor and editing trainer Larry Jordan writes, “New automated editing technology will put lots of us out of work. Probably lots and lots of us. And that’s scary.”
But if we look closer, we find that artificial intelligence in TV and film production isn’t a threat to human creativity; it’s an exciting tool with the potential to push our audiovisual experiences to new heights. While the future automation of some creative tasks will disrupt the job market, it will also present new artistic and business opportunities for those creative professionals who are willing to adapt.
So how can content creators and editors make these new AI technologies work for, not against, them? How can the streaming industry take advantage of the benefits of AI to assist in programming?
In the case of Sunspring, filmmakers provided Benjamin, their AI writer, with a large database of science fiction films and the framework for how to format screenplays. From this, Benjamin produced a script. While the result was a fascinating and weirdly touching melange of science fiction elements, the language of the AI-generated script wasn’t actually coherent, causing screenwriters to “breathe a sigh of relief.”
But what about when AI begins to develop coherent narrative arcs and language? One researcher, Dr. Alexis Kirke, has a potential solution. Kirke, a senior research fellow at Plymouth University, has developed AI tech that, like Benjamin, draws on a database of films to piece together character and plot elements. But Kirke’s algorithm, called Zenman, works with the human screenwriter, suggesting character and narrative elements and helping the writer keep track of complex plot elements. And Kirke hasn’t just used Zenman as an AI assistant; he’s also collaborated with it to pioneer original multimedia performances.
Artificial intelligence promises to enhance human content creation in the realm of editing, as well. 20th Century Fox recently made a splash when they hired IBM’s AI, Watson, to select the most appropriately dramatic scenes for a human editor to include in a trailer for the 2016 movie Morgan. Meanwhile, in a recent paper, Stanford and Adobe researchers revealed artificial intelligence with the power to organize footage and automatically edit it according to certain style paradigms indicated by the human editor.
While both experiments have caused a splash, there’s no need to fear the future of AI in TV and film. Watson didn’t displace the human editor; it freed the human editor of the tedious task of sorting through hours of footage to find the most dramatic moments. And while the Adobe software simplified the thankless task of cataloging hours of footage, video editors were unimpressed by the creative quality of the AI edits resulting from the software.
Even Jordan, the editor who sounded the alarm bell on the artificial intelligence disruption to the editing industry, writes that AI’s ability to do high-level, emotionally and narratively complex editing will take some time to develop. In the meantime, Jordan writes, content creators will have an opportunity to learn new skills and adapt.
When you consider the audience perspective and creator profit, AI in TV and film just makes sense. Already, we’re seeing how artificial intelligence can make more interactive viewing experiences while helping content creators personalize content. My own company’s Media Xperience Studio service, for example, allows content creators to use data on audience reactions to individual clips to decide on the final cuts for TV and film, allowing more engaging experiences for viewers and more effective placements for advertisers.
On top of that, AI revolutionizes the content programming game. For 80+ years since the birth of television, programming has been less of a science and more of a guessing game, relying on focus groups to determine what content to put on the air and when. It’s been proven time and time again that focus groups don’t work very well, telling NBC, for instance, that Seinfeld would be a flash in the pan. But we’ve yet to come up with anything different.
With AI, we can eliminate the guesswork of TV and film and solve the scale issue as well. Why scale? Because programming a content lineup in the future is not a one-to-many problem. With dynamic scheduling APIs, a streaming content channel can look different for each and every user, creating a multiplicity of personalized channels instead of one monolithic stream. AI will help us to deliver those unique experiences to viewers, all while building brand loyalty and increasing viewing time to drive more revenue.
Artificial intelligence offers content creators the potential to take much of the perspiration out of making TV and film, leaving creators more time to make more compelling shows, more exciting videos and “stickier” advertising content. And if the creative machine-human collaborations of Alexis Kirke, or the odd enchantment of Sunspring, are anything to judge by, AI promises to open up a world of creative possibility.
To take advantage of these possibilities, content professionals need to get out in front of them, not hide from them. By learning to love the machine, content creators can make sure that the story of AI in TV and film is a sci-fi plot that ends with happy human-robot coexistence.
