By Cameron Davis

Massive layoffs, cheapened content, and misinformation on steroids: AI text-to-video is coming for Hollywood, and it certainly feels like a curtain call.

Your brain and AI: Wikimedia Commons

In the past year, the pace of advances in AI-generated video has been rapid. Google’s VEO 2, Sora, Kling, Adobe’s Firefly, and Runway have all launched with promises to create realistic, compelling videos from text prompts. Comparisons between the models’ outcomes, using complex prompts such as “golden retriever running through a brutalist art gallery with modern art on the walls as a single red balloon floats overhead,” reflect staggering leaps in quality and believability.

And yet, there are many dire real-world consequences of AI-generated video: criminals use AI videos of public figures to bolster fraud schemes (per a recent FBI alert). There are also more systemic challenges, such as “trust erosion in institutions and evidence falsification issues” as highlighted by a review from University College London researchers.

But AI text-to-video poses an immediate threat to one American industry above all others: Hollywood. The technology threatens not only the nearly half a million US employees in motion picture and sound recording, but the cinematic art form itself.

The film industry is already flirting heavily with artificially intelligent tools. In 2024, for example, director Brady Corbet used generative AI to improve Hungarian accents in his widely lauded immigrant epic The Brutalist. IFC Films defended its reliance on AI-generated images in its Late Night with the Devil.

These examples are just the beginning of a much broader industry shift.

The MIT Technology Review covered an AI film festival in New York, where the completely AI-generated short films in competition had a universally dark, unnerving quality. Artificially intelligent “director” Benjamin created the similarly creepy and jarring Zone Out in just 48 hours with no human involvement. Rather than delivering the satisfying spooks of a traditional horror movie, the warped hands, decaying animal corpses, and ghostly scores in these pictures were the product of something distinctively non-human.

Again, AI-generated video comes with many other risks, too. Advocates dismiss these concerns under the banner of “democratization,” a standard argument in favor of disruptive new technologies in the 21st century. But three aspects of AI video’s relationship with the film industry are particularly troubling.

Winners of the First AI Film Festival – Showcasing the Latest Tools in AI Filmmaking

First, these developments risk alienating audiences from the craft and authenticity that underpin quality art. A popular social media account recently posted a video of Davy Jones, the tentacled villain from Pirates of the Caribbean: Dead Man’s Chest, lamenting the “backwards progression” of computer-generated imagery since the film’s 2006 release. With 46 million views, thousands of commenters debated the issue, but a common thread emerged: effects rooted in reality remain more compelling. Before AI, Jones’ face was textured using scans of real coffee-stained Styrofoam cups, and motion capture technology anchored the CGI character in actor Bill Nighy’s performance.

AI-generated video, however, remains based only in the conceptual; its production is completely unreal. Moving towards AI in lieu of practical effects will further unmoor the industry and further degrade the quality of what is already in cinemas.

Second, AI’s financial impacts arrive at a particularly vulnerable moment for Hollywood. Studios face an impending wave of mergers and acquisitions, streaming competition (with Netflix dominating but rivals growing), and lingering financial fragility five years after COVID-19 devastated box office revenues. These pressures drive cost-cutting and lower-quality results; the adoption of AI text-to-video could accelerate this race to the bottom. Why hire graphic designers and video editors when AI can perform similar tasks for a fraction of the cost? While the film industry’s job losses may pale in comparison to the estimated 300 million jobs that experts say AI will impact globally by 2030, the pressures will be especially acute in an already struggling sector.

Third, and perhaps most concerning, AI-generated video often feels unsettling or even malicious. There appear to be a variety of reasons for that. For example, the aforementioned creepiness of the video works AI creates. Maybe it’s the way that AI’s training on existing major films could expand the dominance of franchises and continue to quash independent storytelling. Plus, AI’s creator-less nature (and relative un-traceability of its creative decisions) makes it prone to murky, irresponsible use. In addition, AI was a central flashpoint for months of union striking in 2023. The final agreement reached by SAG-AFTRA was hardly a comforting assurance that studios wouldn’t pick up AI tools as soon as the agreement expires next year. (Indeed, actors are already being pressured to sign away rights to their digital likenesses, as AI video improves leaps and bounds from when the union contract was inked.)

“AI Generated Videos Just Changed Forever”. Photo: You Tube

But the most troubling aspect of AI-generated movies is the capitalist impulse to use them to placate passive audiences. Real film, like the best art, challenges viewers-questioning the status quo, exploring new perspectives, and probing human dilemmas. By contrast, AI text-to-video often does the opposite.

TCL, the Chinese company that’s the second-largest TV manufacturer in the world, recently announced a slate of AI films as a way to “offer a lean-back binge-watching experience” funded by advertisers. They include the nonsensical and, frankly, hideous Next Stop Paris, whose trailer was released last year. These will be perfect for background noise in a busy household or barbershop, supporting ad revenue to a select few tech companies while stripping away passive royalties to real artists. (That doesn’t even begin to address the geopolitical implications of AI-generated movies; it can be no coincidence that one of the biggest recent viral clips came from fellow Chinese firm Tencent’s generator.)

Because they are unlikely to compete with big-picture releases, AI movies will also target society’s most vulnerable: children. Online, they already do, with AI-generated animated children’s short films racking up millions of views on YouTube. Often full of misinformation or made-up words, these videos are similar to TCL’s passive content – mindless cash-grabs at the expense of real artists and audiences alike.

Massive layoffs, cheapened content, and misinformation on steroids: AI text-to-video is coming for Hollywood, and it certainly feels like a curtain call.

Cameron Davis is a graduate student pursuing a dual degree between MIT Sloan and Harvard Kennedy School. His writing on film has appeared in The Tech, the FASPE Journal, and his review site Movies, Now More Than Ever. He lives in Cambridge.