In an era where one can jot-down words in apps and websites, and images — still or animated, with the strokes of a renaissance painter, the style of a popular anime or feature film, or the realistic guise of a human being — are rendered in seconds, Ted Sarandos, Netflix’s Co-CEO, says that he still has “more faith in humans.” “I don’t believe that an AI program is going to write a better screenplay than a great writer, or is going to replace a great performance, or that we won’t be able to tell the difference,” he told The New York Times in May. Enter then Sudowrite, an AI writing tool backed by the minor investment of screenwriter John August (Charlie and the Chocolate Factory, Aladdin) that, for the price of $22 a month, can help you write a screenplay — words, actions, perhaps maybe thoughts and all. “Blank page, begone” reads one of the headlines on the website. The software helps you “brainstorm” (ie generate) ideas out of thin air, “write” automatically with a selected tone, “describe” the text by reworking it from the sense of sight, smell, taste, sound, touch or metaphor (yes, these “senses” are options one can select to write the words), rephrase the text to be more “descriptive” (one can select the options to “show not tell”, or to “be more intense”) and, technically, teach it to write it in your style. Much like a melting chocolate bar in the hands of a five-year-old, the words “artificial” and “intelligence” have smeared and seeped into most aspects of filmmaking. But what do they mean for the future of films as an art? If, that is, you have a style — or if you even have the ability to write. Given how fast AI is progressing, the last part is an optional requirement. Like a melting chocolate bar in the hands of a five-year-old, the words “artificial” and “intelligence” have smeared and seeped into most aspects of filmmaking. The words, when bundled together, however, mostly represent a misnomer. Unlike scientific and medical research, where the actual prowess of AI is explored — meaning: the use of “intelligence” is used the way it is meant to — the entertainment industry makes do with a watered-down aspect of the technology called “generative AI”, ie images conjured up by descriptive words. In a nutshell, generative AI — popularised by programmes such as Adobe’s Firefly, Dall-E, Midjourney, Stable Diffusion, Flux — is trained on hundreds of millions of reference images, breaking down and cataloguing specific details, be they facial characteristics, colours, hue, art and artist style, painting mediums,brush strokes, types of photography, clothes, or era, amongst a myriad other descriptors. The details from these “text prompts” then fabricate an image on one’s whims. One can train these data sets to specificall reference anything from tractor designs, to the artistic style of Dutch painter Johannes Vermeer, to the face of Nicolas Cage (all three examples can be found at CivitAI, a free generative AI platform), and one can see how easily the low-bars of the hurdles of ethical use can be hopped over. This new reality “really terrifies” filmmakers such as The Daniels (Daniel Kwan and Daniel Scheinert), who made the Oscar winning film Everything Everywhere All at Once — a VFX-laden film, produced at a fraction of a Hollywood film’s cost, exclusively through their human skill. The technology also scares and exasperates actor Nicolas Cage. “They’re just going to steal my body and do whatever they want with it via digital AI... [God] I’m terrified of that!” he said in an interview with The New Yorker this past July. “It makes me wonder, you know, where will the truth of the artists end up? Is it going to be replaced? Is it going to be transmogrified? Where’s the heartbeat going to be? I mean, what are you going to do with my body and my face when I’m dead? I don’t want you to do anything with it!” Cage has been spooked by the use of his likeness at the climax of The Flash — a nod to his unrealised casting as Superman in the Tim Burton-directed film that never left the storyboard and costume design phases. The process he criticises is called EBDR — Employment-Based Digital Replica — and, according to a deal struck with Alliance of Motion Pictures and Television Production (ie the collective of producers) by the actors’ guild SAG-AFTRA, it compensates actors for the use of their digital likenesses. Two months after The New Yorker article came out, Cage, attending the 25th Newport Beach Film Festival, urged artists to be careful. “The studios want [the EBDR technology] so that they can change your face after you’ve already shot it — they can change your voice, they can change your line deliveries, they can change your body language, they can change your performance,” Cage cautioned. “Film performance, to me, is very much a handmade, organic, from-scratch process. It’s from the heart, it’s from the imagination, it’s from thoughts and detail and thinking and honing and preparing.” Cage’s concerns are legitimate. Specialised services such as TrueSync and DeepEditor can manipulate the gamut of a performance, right down to the movement of an actor’s lips, to seamlessly make new scenes or dub the production in another language; the results, evident from the reels available at the website from commercials and films, are hard to nit-pick — if they can be noticed at all. SAG-AFTRA has been fighting to put an effective system in place for actors, pushing and supporting bills in legislation, striking against giant video-game companies Activision, Epic Games, Electronic Arts, Insomniac and WB Games, and inking deals with Ethovox — the only voice AI company owned and managed by voice actors. Some of these use cases, however, appear to be experimental fads. This June, Al Michaels, a noted veteran play-by-play sportscaster, allowed his digital likeness to deliver “AI enhanced” daily recaps of the Olympic Games in Paris on Peacock, NBC Universal’s streaming service. A month later, in August, Angela Murray, a contestant on the 26th season of the reality show Big Brother, was shocked to see a very real virtual doppelganger of herself in the show — an “AI instigator”, designed to create chaos by spreading misinformation in her guise. While these instances can be surmised as marketing gimmicks, the technical end of the filmmaking spectrum introduces bigger problems. A study by the Animation Guild, featured in The Hollywood Reporter (The Hollywood Jobs Most at Risk From AI), deduces that over 200,000 jobs for sound engineers, voice actors, concept artists, visual effect artists, 3D modellers and animators, will be adversely affected by the “ease of use” AI offers. Services such as Meshy (used by giants like Sega) and Nvidia’s Edify 3D and Nim create 3D assets — be they fantasy creatures, mediaeval carts or phone booths — in minutes, rather than days, robbing jobs from specialists with years of trainings in an industry that is already under duress. For years, studios have been pitting visual effects companies against each other in bidding wars to secure jobs at the lowest cost. To survive, individual artists have to deliver multiple iterations of complex visual effects quickly, in the least amount of time and pay. Recent deals, like the one where film studio Lionsgate has allowed AI company Runway access to their entire film library to train AI for VFX work, makes matters worse. In the above-mentioned Hollywood Reporter article, Cameron Scott Davis, a concept artist, says that most positions for his skill set have already dried up. One of his clients, a Los Angeles-based advertising company, stopped hiring him after adopting AI tools that would “produce hundreds of iterations of concepts and illustrations in minutes.” Shortly thereafter, a game studio he went to for an interview admitted to not having any concept artists on their roster. The studio relied on Midjourney — a generative AI service available to the public — to create creative concepts by just typing out text prompts. Just a few days ago, Runway ML, a popular generative AI service available for free to the public — with an unlimited premium offered for a mere $76 monthly — unleashed Act-One, one of the terrifying tools that scares the daylights out of Cage. Act-One can make any character one wishes via text prompts or videos of one’s performance, and move cameras like a real cinematographer. Other service providers — Luma Labs’ Dream Machine, Pika, Minimax AI — are catching up. “AI is not going to take your job. The person who uses AI well might take your job,” Sarandos had said in the NYT article — and, technically, he is right. Apart from concerns of ethical restrictions — actor Morgan Freeman was deepfaked, and explicit images were made of Taylor Swift and Jenna Ortega, leading the young actress to quit the social media platform X (formerly Twitter) — one can find and use these technologies for free, and without restrictions. Platforms such as Stable Diffusion, now backed by Avatar director James Cameron’s support, can be downloaded to one’s computer, freeing one to make any image without consent. In an age where anyone can mimic good-enough “art” by simply typing words on WhatsApp, Facebook, Bing and Google (yes, they all have that ability now), how does the art and the artist survive? Even though the results are sporadically janky in quality right now — and that will soon change — is it a replacement of a person’s talent, be it in the arts, performance, design or modelling? Can prompting images into existence really make one an artist? And, as Cage says in his misgivings, what about emotions that lie at the heart of the art? With AI seemingly poised to take over, aiding the wholesale-manufacturing approach of entertainment in the streaming age, one wonders whether advances in some sectors of technology are really worth it — especially when the human soul is at stake.