The Robot Holds the Remote

Long gone are the days when families gathered on the slightly battered couch, huddled together after a hard day at school or work, enjoying a good run of evening television. This tableau – often portrayed on Channel 4’s Gogglebox as the way that everyone must consume television – has long been fragmented. Netflix, Amazon Prime, on-demand TV, and personalised devices have blown family-television time to smithereens. But just as we’ve finally got used to the idea that watching different shows and not fighting over the remote control does not mean that we’re all from broken families, a scary-sounding technology is hot on our heels.

The key is scary-sounding, because all AI TV screams to me is ‘futile’. I’d switched on the actual box in my living room for the first time in months, blowing dust and dead insects off it, enjoying the masterpiece that is Black Earth Rising, when an unnatural cyber-human voice warned me that they would ‘take control of my viewing’. Knowing that the idea of anyone seizing control of anyone else’s anything would put that anyone on edge, the monotonous voice instructed me to relax: ‘everything’s going to be fine.’

AI…is sifting through the BBC Archives and unearthing hidden gems after learning about the kinds of programmes that prove most popular… This would evidently cost a human being an awful lot of time[.]

Emulating a horror film, the cyber-human control freak as its protagonist, the commercial essentially implied that all television was now going to be artificially generated. I could kiss beginning and end credits goodbye (freeing up a whole minute of my life for more productive, high-efficiency activity, of course) because a robot would do all the work. The creepy dialogue fulfilled every stereotype of the Robot that is circulating in popular consciousness today – and, in doing so, successfully grabbed people’s attention.

AI TV, the brainchild of the BBC’s Research and Development arm and BBC Four, the millennial arm, isn’t really doing what you might expect it to be. Alongside two nights of scheduled programming and demystification of Artificial Intelligence, AI TV in its current form is sifting through the BBC Archives and unearthing hidden gems after learning about the kinds of programmes that prove most popular amongst audiences. This would evidently cost a human being an awful lot of time, not to mention their eyesight and possibly also their hearing. So we let the machines do it.

How the success of this venture is going to be judged is anybody’s guess. The way that TV is produced, in its casting, production value, cinematography and narrative structure, has changed dramatically since the foundation of the BBC, so what is to say that archived TV shows would even appeal to today’s television audiences? It feels as though the BBC is jumping on the bandwagon of AI-everything, trying to reposition itself as a technologically-advanced multimedia platform rather than the legacy television channel that it is.

The very nature of AI is that it is predictive, and it makes its predictions based on previously collected data. With that in mind, is AI really capable of making any predictions about changes and shifts in culture or preference that cannot necessarily be captured in a dataset? Therefore, is AI really able to predict what the next big thing on TV is going to be? Five years ago, Scandinavian crime dramas and thrillers suddenly appeared on the international television scene, their hushed tones and uncomfortable framing making waves with global audiences. The ‘formula’ used to compose a Scandi-drama of the sort was completely unlike that of any other successful crime drama or thriller, and it became a sub-genre in its own right. AI, looking at historical data as opposed to future trends, would probably never have made that prediction. So if the BBC is hoping it can sit back with its feet up now that AI in is in the mix, they’re sadly going to have to keep the creative juices flowing.

There’s a wider risk here in that AI and automation will soon become branding exercises, rather than offering any tangible benefit

That said, AI’s predictive abilities are going beyond simple datasets and singular values now, finding structure, logic, and meaning in unstructured data. So, in the context of the television industry, AI TV is really making waves in the process of editing. The technology is also being employed to learn about, or teach itself about, narrative structures, what makes up a TV scene, and whether a scene is ‘high’ or ‘low’ energy; all this learning then feeds into the AI creating and composing a sequence of scenes into a coherent narrative. As jaw-droppingly cool as this may be, it does throw some of our AI-related assumptions into the air. We’re told time and time again that AI is a threat only to ‘mundane’ and ‘repetitive’ jobs, and that the way to remain above this new wave of technology is to stay creative and forward-thinking. Although TV editing has elements of mundanity and repetition, it is inherently a creative, intuitive process, even more so as new styles of cinematography emerge. ‘Teaching’ a robot a process that is currently being turned on its head might then equate to television producers shooting themselves in the feet.

TV is cool, and getting cooler, without needing the added ‘wow factor’ injection of AI. TV constantly challenges its own notions of acceptability and imagination. It is experimental and future-facing in a way that AI can’t capture, and the two don’t marry together well enough for that exact reason.

There’s a wider risk here in that AI and automation will soon become branding exercises, rather than offering any tangible benefit in the field that they attach themselves to. In the case of the BBC, their toe-dip into AI TV has been met with neither hoots nor boos, because ultimately, the viewer is still very much in control of what they are watching, and hasn’t surrendered any control to the cyber-human on the screen. Whether it’s TV not being ready for AI, or AI not being ready for TV, we will likely continue leaving shows half-finished, zapping away from, or through, adverts, and in reality, probably not watching very much actual TV at all.