Hear Me Out: AI-generated movies will never be interesting
“AI is coming for your job,” bark the doom-mongering Twitter users, seemingly gaining schadenfreude from the steady automation of other people’s livelihoods. Indeed, AI will likely be coming for many jobs, having already taken over from supermarket cashiers, and one day AI may even knock on our doors, coming for our heads, but until then, can we stop pretending that it is any good at making movies?
Computers are certainly capable of creating fascinating visual landscapes, just think back to the spectacular maze of pipes that the old Windows screensaver used to make back in the early 2000s, but so much more goes into making a movie than just aesthetic dazzlement. Gazing at such screensavers, with one’s infantile mouth drooling with fascination, brings mere retinal stimulation, with nothing going on beneath the surface other than the incomprehensible flutter of zeroes, ones and twos.
It would be foolish not to recognise the advancements in such a field, however, with computers and artificial intelligence having grown exponentially smarter in the decades since the dawn of the new millennium. Such technology can store vast amounts of data, allowing them to create grand vistas, realistic human models and much, much more, leading some filmmakers such as Joe Russo of Marvel’s Avengers to predict that AI will be able to create competent movies of its own within “two years”.
These contemporary ruminations have led some people to believe that cinema is doomed, believing screenwriters, editors, lighting experts and set designers should pack up their things and search for new careers. Endless AI creations have followed suit, with people prompting the software to create endless ‘spectacular’ cinematic scenarios, with one online user releasing a fake trailer for a Star Wars movie as if Wes Anderson directed it.
On a very surface level, the AI does a good job pin-pointing the iconography of an Anderson flick like 2012’s Moonrise Kingdom, using centralised symmetrical imagery, pastel colours, kooky humour, and even a narrator that sounds a whole lot like Bob Balaban. Still, any film student half-watching a Wes Anderson movie would be able to distinguish the same features, it takes a lot more to realise the intention behind his cinematic style, and more so the sincerity of the emotions he translates through his screenplays.
AI is incredibly efficient at copying and recreating visual styles, making it a useful tool in assisting the creation of vacuous blockbusters but rather useless in the field of authentic, independent cinema. Essentially only good for visual recreation, AI does not have the processes to feel and understand the importance of mise-en-scène, lighting, blocking and direction.
It is the expert knowledge of industry professionals that adds meaning to such aforementioned elements, just like it’s the screenwriter and director who add purpose and significance to the story on-screen. Without that human intention, we may as well be looking at a series of interweaving pipes on a Windows screensaver whilst Alexa reads us a story based on a boring series of algorithms and codes.
The artist gives meaning to their art, but when there is no face, no personality and no intention behind the work, there is nothing to engage the viewer, well apart from poorly rendered hands that look like fists of cooked spaghetti and faces straight from the uncanny valley that ripple like beige jelly.
