One of the areas I have been exploring this past year is ‘computational creativity’. That is using machines that create written words and images and video. One of the subsets of computational creativity is personalization. We have been working with companies and startups that allow brands to create video that is personalized depending on who is watching – a skill that is called dynamic social video. So if, for example, you are watching an ad on Facebook and you are interested in travel, yoga and cinema, then you will get a different creative. It will be the same message, same narrative, same call-to-action that everyone else gets, but the content will be adapted to your interests.
This is totally possible now and smart brands are using dynamic social video to make their content more targeted. But if you extrapolate out from there it is not beyond the realms of possibility that, at some point, all TV programming could become individualized versions of content. For example, a new character in Stranger Things could have a different accent, be a different gender or a different race depending on who is watching it.
So how is that possible? Well, it is down to something called a generative adversarial network. There are lots of uses for generative adversarial networks, but one of the things MIT has been using it for is to create video from photographs by using adversarial networks to work out what the next frame would be. So, if you have got the combination of audience data then we can generatively create content on the fly to suit your interests and your background. So the future of TV is all of us seeing the same stories, but getting served personalized versions of that content to suit who we are. In the same way that currently you release a film in a lot of different languages, it is an extension of that.
The BBC experimented with it a few years ago, where it had a camera on top of the TV and, depending on how you responded to the film, it would give you a different storyline. We are now experimenting with that kind of content for in-store. Some might see a danger in TV or film being over-personalized if they only ever see versions of themselves, but it doesn’t have to be that way. It uses who you are as a data point in order to provoke a certain reaction in you.
For example, if the viewer is a white, middle-class male, the director might say we should show predominantly non-white people in this scene. It is not about showing people who are like ourselves, it gives the creator the opportunity to frame a different experience based on that way that person identifies themselves on social or how they appear. It runs the risk of being weird and freaky, but, like any technology if it is used sensibly, elegantly and respectfully, it should produce something that is remarkable. It is the start of a whole new era of creativity where we use human insight and machine creativity together.
The future of TV is one where we use human insight and machine creativity together to produce personalized content.
Tom Ollerton is the innovation director for We Are Social.
The Drum will be hosting the Future of TV event in New York on 8 May, for more information on how to attend visit the following website or to find out more about how TV will develop, purchase The Drum’s dedicated Future of TV issue online.