If you’ve followed the news recently, you’ll be aware that there are ongoing strikes in Hollywood, with writers and actors forming picket lines for several weeks with no end in sight. This has disrupted many productions, including cinema, television, and chat shows. Of course, part of the strike action is down to remuneration and general terms, but the most fascinating aspect is the protest against artificial intelligence and its role in creating film.
For those unaware of the details, it’s worth noting that the strikers aren’t simply against movie studios using AI in and of itself. It is a little more nuanced than that. AI is inevitable, and protestors against using the technology understand it will likely disrupt their industry. They are not alone; everyone from lawyers to software engineers to graphic designers could soon see their vocations impacted by the technology.
AI is Inevitable in Cinema
However, there is a saying among AI experts that broadly says, “AI won’t take your job; someone using AI will.”. This is important for understanding the role of AI in broader society and, indeed, its relationship with cinema. You see, despite the jaw-dropping ability for AI to “create,” it can only do so within parameters. It cannot reason to provide original information, only interpret it. And despite proponents saying otherwise, it lacks the spark of ingenuity that comes with true creation.
The above is at the heart of the Hollywood strikes. Writers, for instance, have said that they do not want AI to be trained in their work. And if AI is not given the data to learn, then it cannot create. This is a crucial point, and it has gone largely unremarked upon in the mainstream coverage of the issue. In short, an AI could make a passable Martin McDonagh film script, but it would have to train on the works of the director. The writers (rightly) ask why they should allow their work to train their potential replacements.
AI essentially trains on internet data. But consider this premise: How would an AI train to create a film about modern Ireland? News clippings, blogs, and other media would help, as would the vast amount of information on social media. But other, widely unrelated data points crop up time and time again: The Celtic symbolism that you’ll see in Irish gift shops and Irish-themed slot games, the outdated stereotype toward alcohol, and many other things that are no longer representative of modern Ireland.
The Creative Spark is Lacking
Writers both create and interpret. In the same way that Roddy Doyle provided an insight – but his insight – into parts of working-class Irish life in the 80s and 90s, contemporary writers are being forged in the cauldron of the Celtic Tiger, the housing crisis, the question of reunification, and so on. The stories we tell about ourselves are fluid, and AI is – at least not yet – able to keep up with that fluidity. Moreover, AI considers the interpretations valid: Doyle and McDonagh see Ireland in a certain way, but it is characteristic of their worldview.
To momentarily play the devil’s advocate, one might say that all humans are products of their environment and legacy. Oscar Wilde was influenced by John Ruskin and Walter Pater, among others, and he was built on the works of the writers, artists, and thinkers he admired. An AI can work in broadly the same way. You can feed it on the films of Lenny Abrahamson, and it’ll produce scripts and ideas that are characteristically like the (brilliant) director, but it cannot replace the spark of ingenuity that Abrahamson is capable of.
And that, perhaps, is the point. Those striking in Hollywood understand that AI will be used to replicate, not create. The difference between the two concepts is essential. If movie bosses get their way, AI will be trained on everything that has come before; it might even produce quality entertainment. But it is not capable of the sudden leaps in creativity that act to amaze us with cinema. If that occurs, it won’t only the writers who lose out.
3.4k
20k
1.5k
1.3k