FILE PHOTO: Sunspring, a short science fiction movie written entirely by AI and performed by human actors, was released in 2016.

In the age of artificial intelligence, the rise of the algorithmic film industry is reshaping cinema. Driven by algorithmic prediction, production, and consumption, this new model is defined by a logic of affirmation in prediction, technological logic in production, and traffic-driven logic in consumption. It intervenes throughout the entire filmmaking process—from planning, shooting, and production to distribution, feedback, and evaluation. Films produced within this framework are known as algorithmic films—works that bear clear traces of algorithmic processing or are even entirely generated by intelligent algorithms. This emerging industry not only updates and, to some extent, disrupts traditional filmmaking, but also spurs aesthetic shifts and a fundamental revolution in cinema. 

Logic characteristics 

Traditional film creation, from conceptualization and scriptwriting to screenplay development, often requires years or even decades of revisions and refinements, testing both the intellect and endurance of creators. In this long and arduous process, ideals and reality constantly constrain each other, making the filmmaking journey—both on and off set—full of surprises, setbacks, and turning points. Admittedly, various real-world obstacles can hinder film production, but it is precisely these elements of negation that lend cinema its modern appeal. 

By contrast, the algorithmic film industry, with its calculations and predictions, aims to eliminate negation. Algorithmic technologies provide creators with greater means of self-affirmation. During conceptualization, algorithms can analyze keywords and mine vast databases, enabling creators to swiftly grasp and capitalize on trending topics, thereby significantly shortening the planning cycle. In scriptwriting, algorithm-enhanced evaluation systems assess script quality and offer recommendations, while AIGC (AI-generated content) technology can even produce drafts that, though not fully polished, serve as usable blueprints. In the screenplay development phase, virtual previsualization technologies translate creative imagination into concrete visual simulations, offering creators intuitive feedback for revision and refinement, and enhancing the feasibility of subsequent filming. 

Under the precise forecasts of algorithms, each stage of the film industry becomes more certain and transparent—like a darkroom flooded with light, stripping away its mystery. In the celluloid era, creators worked much like photographers in a darkroom, only discovering the final image upon development. Even in the digital age, while monitors and other devices facilitated collaborative filmmaking, creators on set still operated within a fragmented reality—the darkroom lingered. It is only in the era of artificial intelligence that the algorithmic film industry, empowered by technological capabilities, has finally dismantled this creative darkroom. 

Filmmaking has become a process of translating algorithmic predictions into visual form, with creators no longer confronting the uncertainties of darkness. Instead, they work in brightly lit environments, guided by risk assessments and the projections of virtual previsualization. 

The algorithmic film industry avoids negation across production, screenwriting, and cinematography, constantly pursuing affirmation and transparency in film creation. The logic of affirmation underpinning algorithmic prediction not only enhances the efficiency of the film industry’s machinery but is poised to increasingly shape both creativity and practice. As precise forecasting and high-accuracy previsualization advance, algorithmic filmmaking will continue gaining momentum. However, critical reflection remains essential. Creators must remain alert to the risks of an overreliance on affirmation and transparency, lest negation and diversity vanish within algorithmic calculations. 

The large-scale application of algorithmic technology in the film industry is driving a significant aesthetic shift. Leveraging digital technologies such as virtual reality (VR) and augmented reality (AR), algorithmic films have expanded the realm of “hyper-real” imagery, influencing traditional cinematic aesthetics. It is foreseeable that AI algorithms will continue to exert transformative power in the film industry, enhancing its artistic ambitions through technological innovation. Over time, the industry will refine its ability to generate realistic human gait, movement, and micro-expressions, while integrating AR, MR, XR, and VR more deeply with real-world imagery, advancing the fusion of technology and art. 

Traffic is key 

In the traditional film industry, box office performance has long been the primary metric of success. In the algorithmic film industry, however, traffic has emerged as the key criterion for evaluating consumption. 

The first defining feature is data-driven precision consumption. Relying on big data technologies, the algorithmic film industry develops precise commercial strategies from the early planning stages. Through constant audience monitoring, it readily grasps viewer preferences, swiftly identifies genres, targets niche markets, and tailors creative efforts accordingly. Once production is complete, big data further enhances algorithmic recommendations, ensuring targeted distribution and guiding audience consumption. 

At first glance, these recommendations appear to offer personalized attention to each viewer. In reality, they reduce the need for audiences to actively seek out content, turning them into passive consumers of cultural products. Smart recommendation algorithms monitor users’ browsing behavior within their personal environments, assess their preferences, and push similar content accordingly. Alternatively, through collaborative filtering, user behaviors are transmitted to others, aggregating individual tastes into dominant trends.  

Next is platform-driven segmented consumption. If data-driven precision targets individuals, platform-driven segmented consumption gathers and categorizes target audiences on a larger scale. In the internet era, online communities—Weibo super topics, Douban groups, and the like—have made moviegoers more connected than ever before. Audiences are now able to find like-minded peers with unprecedented ease. 

The algorithmic film industry has also targeted the commercial benefits of community-based viewting. By analyzing the popularity of celebrities on platforms like Douyin and Weibo through big data, film producers can directly assess and compare the traffic value of trending actors. 

Lastly, there is mass consumption shaped by public opinion. Unlike the traditional film industry, where box office figures and critical reviews serve as the main evaluative criteria, the algorithmic film industry, with its heavy reliance on traffic promotion, places greater emphasis on managing public sentiment. Generally, a film’s publicity strategy and public opinion campaign are prepared in advance. With big data support, the algorithmic film industry can deftly manage sudden public opinion crises, consciously steering sentiment in a direction favorable to the film’s promotion. 

Critical reflections 

To guide the healthy development of the algorithmic film industry and foster the transformation and upgrading of traditional filmmaking, it is essential to critically reflect on the challenges exposed by the rise of this new model. 

Foremost is the issue of rupture. The algorithmic film industry, shaped by its triple logic, highlights rupture on three levels: between the author and algorithm, between algorithmic and real imagery, and between algorithmic and real audiences. 

First is the rupture between algorithmic authorship and human authorship—a divide between algorithmic creation and artistic creation. Here, the role of the human creator is displaced by algorithms, resulting in a fundamental conflict regarding the nature of creation. 

Second is the rupture between algorithmic imagery and real imagery. This refers to the divide between the algorithmically constructed world and reality itself. While algorithms can simulate and predict imagery, they cannot fully replicate the authenticity and depth of real-world experience. 

Third is the rupture between algorithmic audiences and real audiences. The algorithmic audience is data-driven and regulated. The algorithmic film industry no longer views audiences as active, dynamic interpreters but rather as tagged entities, ready to be decoded and manipulated. Today’s cultural industry uses increasingly precise means to calculate individual preferences and generate group effects desired by capital interests. In an era where viewers’ preferences are meticulously collected and categorized, it becomes more difficult to cultivate unique, independent tastes. 

Next is the issue of hegemony. the algorithmic film industry exhibits two key features when exercising power and shaping recognition. 

First, a shift from artistic recognition to technological recognition. In appreciating algorithmic films, technological spectacle gradually eclipses artistry, forming a visual hegemony over audiences. The immersive “shock” of 3D films and high frame rates draws attention to technical prowess, overshadowing the intricate emotional resonance traditionally conveyed through art. 

Second, a shift from human recognition to non-human recognition. In the traditional film industry, films garnered artistic recognition through the careful direction of filmmakers, compelling scripts, and superb acting. Ultimately, it was human creators who persuaded audiences. In contrast, the algorithmic film industry replaces these human elements with algorithmic control—scripts, direction, and performances can all be generated and assembled automatically. Film could become a manipulation of the audience by non-human algorithms, calculated and curated. 

Finally, there is the issue of subjectivity. In the traditional film industry, creation revolved around the producer or director, with human subjectivity at its core. Humans used and coordinated machines to make films possible. The algorithmic film industry, by contrast, has disrupted this stable creative structure, intertwining the interests of creators, algorithm developers, and algorithms themselves, resulting in complex subjectivity challenges that cannot be ignored. In this light, industry practitioners should adopt a dialectical approach, cautiously analyzing the algorithmic film industry’s development. 

First, one must not overlook the industry’s irreversible trajectory or deny its potential by clinging too tightly to human subjectivity. As artificial intelligence and big data technologies advance, algorithmic filmmaking will inevitably produce works with more dazzling special effects, more dreamlike scenes, and even self-generated audio-visual compositions and coherent storylines. This technological progress responds to creative ambitions, market demands, and public aesthetic preferences. It would be a mistake to neglect its breakthroughs and transformations in the essence of film simply because its artificial nature is less apparent. 

Second, one must avoid an overly technological perspective and ensure that the subjectivity of human creators is not eclipsed. For now, the algorithmic film industry has not yet granted algorithms the capacity for self-awareness—artificial intelligence and big data cannot independently produce films with self-reflective thought. Both AI and humans are capable of rational thinking, but humans also possess super-rational capacities beyond the reach of machines. These unique abilities give traditional creators the potential to produce works imbued with “distance” and depth—something algorithms cannot achieve on their own. 

Liu Jun and Jia Yixing are from the Art and Culture Research Center at Communication University of China. 

Edited by WANG YOURAN