As a creator who transitioned from directing to producing, I have learned that the need for creative autonomy frequently clashes with the deeply collaborative nature of filmmaking. Moving to the producer’s chair allows you to stamp your vision across multiple projects, guiding a team without suffering the microscopic control of studio executives. You learn to see the industry from a macro perspective. Looking at the landscape in October 2024, that macro view is undeniably bleak.
Mainstream cinema has become boring, corrupted, and artistically imploded. In previous decades, the industry was violently refreshed by bold movements. We had the French New Wave, the grit of 1970s New Hollywood, the razor-sharp, twist-heavy plots of the 1990s independent boom, and the raw digital provocation of Dogme 95. Today, aside from the rare miracles pulled off by boutique distributors like A24 operating as a modern-day Miramax to champion auteur-driven risks, the broader industry remains trapped in a sterile loop. The pandemic accelerated streaming adoption, fatally weakening the commercial viability of films devoid of massive visual spectacle or pre-existing intellectual property. The mid-budget movie is dead. Films like Sleepless in Seattle or Erin Brockovich would never survive theatrical development today. Instead, they are relegated to straight-to-streaming content dumps or stretched into bloated eight-part miniseries. Platforms like Netflix and Apple TV burn billions on algorithmically designed shows that seemingly no one asked for, substituting true cinematic movements with a lukewarm slurry of "content."
In this landscape of creative stagnation, the arrival of generative AI video models feels less like a threat and more like a long overdue earthquake.
We typically imagine AI art as a magic button that instantly produces a flawless masterpiece. When considering AI in a broader sense, people often picture a Skynet-like entity performing its own independent actions. We are far from that stage of superintelligence. Whether an AI possesses free will remains a fascinating debate. According to Frank Martela [https://www.aalto.fi/en/news/ai-meets-the-conditions-for-having-free-will-we-need-to-give-it-a-moral-compass], functional free will requires three core conditions. First, intentional agency means possessing encoded goals and motivational states. Second, alternative possibilities dictate encountering genuine choices rather than executing a single rigid script. Finally, action control requires actions guided by internal decisions rather than purely mechanical causal chains.
Contemporary generative AI operates much more like a slot machine. You press a button and pray to win on the first round. Usually, the process demands countless iterations, heavy retouching, and significant manual human effort. The more you know what you are doing, the more likely you will achieve an outcome close to your original vision. Authors of popular viral videos constantly confirm spending hours on the final mile of fine-tuning. The human artist remains firmly in charge, simply operating a much more sophisticated tool.
Right now, the industry is gripped by a moral panic. OpenAI recently suspended access to its Sora video generation tool following fierce protests from artists, while competitors like Runway, Luma, and Minimax aggressively push the boundaries of text-to-video capability. Hollywood is terrified. Yet, as tech analyst Benedict Evans [https://www.ben-evans.com/benedictevans/2023/8/27/generative-ai-ad-intellectual-property] recently observed, we are actually looking at two completely different art worlds. The first world revolves around the value of human intent. In this arena, the customer purchases the artist's identity, personal narrative, and creative integrity. Here, generative AI poses zero threat. The artwork serves as an extension of a specific human being's lived experience. A machine cannot replicate the cultural status, biography, or intentionality that collectors at high-end auction houses like Christie's pay millions to acquire. The art is just as much about who made it and why as it is about the visual output.
The second world is built on the commodity of visual output. In this commercial sphere, the customer simply wants a functional asset, a specific image, a precise style, or a decorative piece to fulfill a commercial need. The buyer cares absolutely nothing about who made it. Because this buyer prioritizes the final look, speed, and cost, generative AI acts as a direct substitute. In this market, human labor is treated as a basic utility, placing commercial illustrators, graphic designers, and stock photographers at immense financial risk. If you operate in that second world, churning out generic storyboards or predictable streaming filler, you have every right to be terrified.
We must understand that copyright, the shield behind which the studios currently cower, was never designed to protect the soul of the artist. It is a relatively modern concept invented to protect corporate monopolies as a response to the printing of scandalous books and pamphlets [https://en.wikipedia.org/wiki/Copyright]. The Licensing of the Press Act of 1662 and the Statute of Anne in 1710 were enacted in England to protect the business interests of the government-approved Stationers' Company, granting them the right to regulate printed material. Today, copyright exists primarily to protect the business of Mickey Mouse and the likes [https://en.wikipedia.org/wiki/Copyright_Term_Extension_Act].
This brings us to the current legal battlefield. The music software startup Suno is currently accusing Universal, Sony, and Warner of weaponizing copyright lawsuits to stifle AI competition [https://www.axios.com/2024/08/01/ai-music-suno-lawsuit-sony-warner]. The studios claim AI companies are stealing. The AI companies argue that learning is not infringing, asserting that it never has been. From a cinematic perspective, the AI companies possess a valid philosophical point. The history of great cinema is the history of masterful theft: Brian De Palma stole from Alfred Hitchcock; Quentin Tarantino stole from Ringo Lam and Sergio Leone etc. If we celebrate human auteurs for metabolizing the history of cinema into new collages, why do we suddenly cry foul when an algorithm performs the exact same process of pattern recognition?
Analyst Benedict Evans accurately diagnoses this growing collision [https://medium.com/@nextconf/ai-as-new-media-e02d34c49c3b]. He notes a fierce divide between the philosophical view held in many AI circles and the rigid legal status of copyright. Tech developers argue that training by looking is fundamentally a human action. These systems do not operate like Napster; they do not store and reproduce exact copies of stolen training data. Instead, they infer abstract patterns of collective human intelligence. Yet, content creators counter that the sheer industrial scale of this ingestion changes the nature of the act entirely, utilizing human property in an unprecedented way without any new model for permission. Just as the music industry eventually transitioned from unregulated digital piracy into the licensed Spotify ecosystem, Hollywood will inevitably force AI companies into massive commercial licensing partnerships to close this legal vacuum.
Alex Hirsch, the creator of the acclaimed animated series Gravity Falls, noticed a key distinction [https://laist.com/news/ai-hollywood-animation-filmmaking]. He stated that the point of art is a human communicating a feeling to another human, rather than an algorithm amassing a slurry of thoughts other humans have had before to reproduce a generic, predetermined outcome. The difference lies entirely in the human element. The actor Ben Affleck [https://x.com/jorilallo/status/1858211286066073922] recently offered the most articulate and realistic explanation of where Hollywood is heading in the AI era. Craft is knowing how to work, Affleck noted, whereas art is knowing when to stop.
AI is a master of craft. It can generate pixels, mimic lighting, and structure a three-act screenplay. As Wharton professor Ethan Mollick suggests [https://www.oneusefulthing.org/p/automating-creativity], a fluent creator can use AI as a brilliant sparring partner to cure writer's block, rapid-test narrative structures, and iterate workflows at lightning speed. Even in sensitive studio cases where a final manual human review is necessary, rapid A/B testing of prompts through synthetic AI reviews saves massive amounts of time and improves results considerably. However, the high cost and latency limits of running extensive tests with AI models like GPT-4 can still present substantial barriers. AI has no lived experience. It has no taste. It does not know when to stop.
It is no wonder a strong backlash exists among established filmmakers. Hayao Miyazaki declared the lack of human intent in AI-generated animation deeply disturbing. Guillermo del Toro argues that the true value of art derives from the human effort and emotion behind it. Tim Burton has expressed that AI strips away the soul and humanity of an artist's personal style, feeling like a mechanical copycat. We must ask if this is ultimately a dispute about art, or simply a dispute about money. One could argue that, similar to any royalty-based business, those who stand to lose the most financial ground remain the strongest defenders of their own interests. Meanwhile, the upcoming underdogs hacking every possible way to break into those elite circles will argue passionately for freedom of expression and unrestricted creativity. Just as video art and the Xerox machine opened up new experimental avenues in the pop art movement, these new tools will forge their own artistic paths.
This is why legendary screenwriter Paul Schrader [https://www.indiewire.com/news/general-news/paul-schrader-ai-can-mimic-best-filmmakers-1235086328/] remains refreshingly pragmatic about the technology. Schrader argues that the Writers Guild of America does not actually fear AI; it fears not getting paid. He acknowledges that AI can mimic the best filmmakers, but he insists on a simple labor rule. If a WGA member employs AI to break a story, they must be paid as a writer. If a producer uses AI to generate a script, they must hire a WGA member to polish and humanize it, paying them accordingly.
We have survived these technological panics before. Every era introduces a crisis that threatens to destroy the medium. The transition to sound, the invention of television, the VHS revolution, the digital camera, and the CGI explosion all forced the industry to completely reinvent itself. The industry introduced solutions like CinemaScope, 3D, IMAX, and Dolby Surround sound systems specifically to draw people back into theaters, ensuring audiences experienced emotions they could never replicate at home.
Consider the smartphone. Luma AI CEO Amit Jain [https://techcrunch.com/2024/11/05/as-generative-ai-gets-better-what-will-happen-to-artists] recently stated that a technological arc is never complete until it is in the hands of literally every single person. When high-quality cameras were embedded into iPhones, critics warned it would destroy photography. Digital photography is certainly more ubiquitous now, and perhaps fewer people hire professionals for basic events. Yet, no one can seriously argue that the art of photography is somehow less valuable today than it was fifty years ago. There remains a fundamental, undeniable difference in quality between a master cinematographer and a teenager with a smartphone. The same will apply to generative video. When everyone can generate a movie from their bedroom, the sheer abundance of synthetic content will make true, human craftsmanship infinitely more valuable. The artisanal, handmade film will become the ultimate premium product.
However, we face a distinct, unprecedented psychological danger. Stephen Fry [https://stephenfry.substack.com/p/ai-a-means-to-an-end-or-a-means-to] has drawn a definitive red line. The threat is not AI making art; the threat is AI masquerading as a human. Self-disclosure must be mandatory. Whenever an AI is drawing, writing, filming, or composing, it must be glaringly apparent that an algorithm is speaking. Fry argues that any digital pretense should fall foul of international forgery laws, requiring digital watermarks as complex as those used on banknotes.
This fear of the masquerade is backed by recent data. A 2024 study by Simone Grassini and Mika Koivisto [https://www.sciencenorway.no/art-artificial-intelligence/people-liked-ai-art-when-they-thought-it-was-made-by-humans/2337417] revealed a fascinating paradox in audience psychology. People actually like AI-generated art, but only when they falsely believe a human being created it. Furthermore, audiences proved terrible at assessing whether an image was synthetic or human-made. Adobe design instructor Martin Perhiniak highlighted an even darker irony within this landscape. The more distinct and unique an artist’s style happens to be, the more successfully an AI can mimic it. Highly complex, surreal, or detailed human artwork is now routinely dismissed by cynical internet users as AI, even when it is entirely genuine.
We cannot rely on algorithms to police this. Algorithms have no stake in society. We rely on social mechanisms like shame, morality, and the law to combat human biases, but these concepts mean nothing to a neural network. The main barrier remains complexity. Because no one understands exactly how AI works, it is nearly impossible to exert full control over it.
Marshall McLuhan famously declared that the medium is the message. Generative AI will inevitably carve out its own distinct community, creating entirely new forms of media that we cannot yet categorize, just as the advent of the portable video camera birthed the video art movement. The old media will survive, but it will be forced to adapt.
The current collision between Silicon Valley philosophy and traditional Hollywood legality forces us to reevaluate the very definition of creativity. Startups have always operated by releasing a product to find an audience first, waiting for the regulators to catch up later. Uber did it to the taxi industry, and AI is doing it to the studio system. Ultimately, the users win. Jobs will inevitably shift, just as the rise of massive VFX departments drastically reduced the need for practical stuntmen and matte painters. Yet the fundamental economics of Hollywood remain unchanged. Historically, massive commercial film productions always subsidized smaller, experimental ones. A studio produces ten films hoping one massive blockbuster will cover the losses of the rest. Today, streaming giants like Netflix can endure immense financial risks, reflecting a modern consumer taste focused heavily on mobile consumption and live experiences. In this volatile environment, early panic is giving way to pragmatic adaptation. As recent industry reports suggest, AI is currently being utilized as a powerful tool by artists and filmmakers to dramatically enhance their own practices, rather than simply replacing them outright.
If AI can effortlessly generate the slick, soulless, formulaic content that currently clutters our streaming platforms, human filmmakers will be stripped of their ability to coast on mediocrity. They will be forced to return to the one thing an algorithm cannot generate: a genuine human feeling communicated to another human being. In a strange, highly disruptive way, AI might be the exact shock to the system required to save cinema from its own boredom.
Brak komentarzy:
Prześlij komentarz