AI opens new possibilities in documentary filmmaking, from sorting archives to speeding up production. But documentary is not built on technology alone: it depends on the director’s vision, creativity, and ability to shape meaning. Can AI, for instance, make a film like Ordinary Fascism?

Truth ,  and the extent to which a film can encompass it ,  has long been a challenge facing documentary filmmakers. Documentary has always been the subject of debate between those who see it as a faithful transmission of reality as it is, and those who regard it as a creative art form shaped by the director’s vision. Today, the boundaries of truth face an even greater challenge as artificial intelligence has become a heavy presence in the world of filmmaking. With its rapid development and ease of use, AI is no longer confined to tools that merely facilitate production. It has become part of documentary construction itself: image-making, archives, effects, sound, music composition, and, in many cases, even the creation of an entire film using its own tools. 

Before the age of artificial intelligence, many directors ,  especially the pioneers of cinema, whether documentary or fiction ,  confronted themselves with a host of questions about the form of the film, its message, its emotional charge, and its relationship to the human being whose story it tells. Different schools and methods emerged in screenwriting, cinematography and editing, and each director left a distinctive imprint that set him apart from others. Film was never merely an art of entertainment; what mattered just as much was the effect it left on audiences. 

“Ordinary Fascism” 

When Soviet director Mikhail Romm was preparing to make his film Ordinary Fascism in the early 1960s, he faced two principal challenges. On the one hand, he found himself confronted with an immense volume of archival material produced by the Nazi regime, German media, Allied armies and private archives. On the other, he encountered the familiar dilemma facing filmmakers: how to reconstruct the events of a film. 

Romm initially wanted to adopt “montage of attractions” ,  or what is sometimes called “montage of appeal” ,  pioneered by Soviet director Sergei Eisenstein. The method relies on linking separate, contrasting shots to produce an immediate shock and create tension for the viewer, ultimately leading them to an intellectual conclusion or a specific position. Yet Romm considered this style of montage too excessive a means of conveying his idea. 

Some may see it as absurd to begin with children’s drawings and then move to scenes of ordinary life, but this unexpected opening was key to engaging the audience emotionally before plunging into tragedy. Accompanied by a voice-over monologue reflecting human values and the details of everyday life, the shock was not immediate so much as it emerged through the gradual accumulation of images. 

To move beyond the harsh shock effect of “montage of attractions,” Romm found the solution in abandoning linear historical narration, so that the film would not become merely a cinematic history lecture. Instead, he organized and arranged the archival material into large, coherent thematic groups. He ended up with 120 subjects, including wide shots of people shouting in support of the Nazi regime, Hitler’s speeches, running crowds, military parades, corpses, the wounded, and a long list of other homogeneous categories. In this way, the film’s form became clearer to him. 

The total length of this material, shot on film stock, reached 40,000 meters after its initial assembly, then was reduced to 15,000 meters. In the process of editing according to the principle of conflicting juxtaposition between scenes, much of this material, as Romm put it, “isolated itself and disappeared.” 

The Soviet director Mikhail Romm wrote: “Our propaganda will naturally take care to show through images the greatest possible number of atrocities, and it will certainly make sure that the file contains the image of an eagle with claws full of blood or a skull wearing a fascist helmet ,  all this will be there. People are waiting to see what we will begin with. And we begin with a laughing cat drawn on the screen ,  by my grandson, incidentally. When I asked him why the cat was laughing, he replied: because it had eaten a mouse.” (1) 

Some may regard it as pointless to begin with children’s drawings and then move to scenes from ordinary life, but this surprising opening was the key to drawing the audience in emotionally before the descent into tragedy. With a voice-over monologue reflecting human values and the details of daily life, the shock was not direct so much as it was built through the accumulation of images. 

Romm wanted the viewer, after relaxing and laughing, “to accept the second, tragic part, which begins immediately with a very serious discussion of how fascism transforms the human being, and what a person transformed into a fascist is capable of doing.” (2) 

The goal was not merely to make a powerful documentary capable of delivering a shock. It was also to ensure that the viewer would not shut their eyes in horror at what appeared on screen, and would understand that what they were seeing was only part of the past. In this film, the director imposed his vision and philosophy. He spent several years working on it until it reached its final form, placing the viewer within a fluctuating emotional trajectory that compels reflection. 

What does this have to do with artificial intelligence? 

At first glance, the question ,  can artificial intelligence stand in for Mikhail Romm in arriving at such a treatment of an idea that requires a special kind of awareness? ,  may seem obvious on the one hand and superficial on the other, once one delves into the depths of how AI actually “thinks.” 

Another question suggests itself here: could a generation raised on artificial intelligence ,  a generation whose contours we do not yet know ,  produce a director able to plunge into the depths of the human psyche and find solutions, even through AI tools, in a way that can flip through the pages of consciousness like the transparent animation cels once used in cartoon-making, or inhabit his mood and uncertainty amid a multiplicity of choices that shift with every image and every shot he watches? 

Today, believers in artificial intelligence will say that it is easier to use technical tools to compile footage, sort it, classify it and even assess its quality and suitability for screening. A director could save himself long hours of labor and go even further by relying on ready-made editing templates to construct a narrative that confronts the audience with the intended meanings and messages. This may even succeed in creating a film built on contrast and suspense, producing the desired shocks through the power of the footage itself. 

Skeptics, however, will see the use of such tools as a marginalization of the director’s role, an erasure of his imprint, and a constraint on his creativity. It would lessen reliance on the feelings affected by each shot he watches with his own eyes, feelings that allow him to see what others do not and impose on him a daily preoccupation and a continuous, shifting thought process about the form of the film before it reaches its final shape. Watching does not arise from data analysis; it springs from human consciousness and, very often, from personal experience, indeed from the director’s immersion in the subject of his film before he even begins to write the script. What, then, if the film is itself an embodiment of the director’s own suffering? 

In the film Scenes from the Occupation in Gaza, produced by the Palestinian Cinema Group at the Palestinian Research Center in 1973, director Mustafa Abu Ali assembled footage shot by foreign journalists and filmmakers to serve the occupation’s narrative, then re-edited and reassembled it to convey the reality of the daily suffering endured by the people of the Gaza Strip. The film, which writer and critic Bashar Ibrahim considers “one of the most important short documentary films produced by the cinema of the Palestinian revolution” (3),  is the product of a cause the director not only embraced, but understood deeply through the lived suffering of his people. 

Returning to the debate over truth, one might ask: what would Mikhail Romm have done in an era when images and sounds can be generated by artificial intelligence, had this immense archive not been available? How would he have written or imagined the script of his film? 

Writing a documentary script, as American director Barry Hampe says, means “thinking in images”,  in other words, imagining yourself sitting in a movie theater seat, looking at the screen. (4) 

This takes us back to the heart of the discussion and the enduring debate over the director’s role, his fidelity in conveying reality, and the extent to which truth can be trusted in his films. 

The “spirit” of documentary film 

In documentary film, technology is not the essential concern. What matters more is understanding its complex and layered nature, and the director’s ability to produce positions and questions that touch human beings and their varied concerns, ensuring that meaning is not displaced by the spectacle generated by artificial intelligence tools, and that reality is not reinvented at the expense of truth. 

An emphasis on form at the expense of substance, without understanding that each film has its own unique story, with feelings belonging to both its subjects and its director, who leaves part of his soul in it, will produce films emptied of spirit, governed by the machine rather than by the human being. 

Who among us has not watched a film dazzling in its visual beauty and technical sophistication, yet felt that it was “soulless”? It is difficult here to define the nature of that spirit, because it is bound up with the inner life of the human being,  his emotions, sensations and connection with the filmmaker in thought and contemplation. 

Romm said of his film Ordinary Fascism: “The film is constructed as the author’s philosophical meditation. It expands the framework of documentary material and compels one to think about the fate of man and humanity within very deep and profoundly contemporary realms. I wrote the text of the film myself; it was not written in advance. We assembled the film as a silent artistic work, and I commented extemporaneously on large sections, without concern for synchronization and without chasing after stereotypical documentary effects. It was as though I were thinking through the material and inviting the viewer to think with me. It is precisely this method,  the combination of artistic montage saturated with emotional relation and the author’s monologue,  that, in my view, gave the film its distinctiveness.” (5) 

In documentary cinema, the primary concern is not technique. What matters more is grasping its complex and composite nature, and the director’s ability to generate positions and questions that touch human beings and their many concerns, so that meaning is not replaced by the spectacle fashioned by AI tools, and reality is not fabricated at truth’s expense. 

Improvisation stands in sharp contradiction to the logic of artificial intelligence, which operates by producing clear texts and answers or a narrative based on fixed templates and patterns. In such a system, the script is not alive, not subject to change from one moment to the next. For Soviet director Andrei Tarkovsky, the script is “a living, fragile structure, always changing,” and the film “does not take its final form except at the moment the work is complete.” (6) 

Artificial intelligence and the director’s mind 

The 1998 American film The Truman Show tells the fictional story of Truman Burbank, who lives what appears to be an ordinary life before discovering that his entire world, from birth onward,  is nothing more than a television studio for a live-broadcast program, and that everyone around him is an actor. He begins to rebel against this manufactured reality. The film reaches its climax when he tries to escape, and the director, hoping to persuade him to return, tells him: “I know you better than you know yourself.” Truman replies: “You never had a camera in my head.” 

This exchange can be applied to the experience of artificial intelligence: can it enter the director’s mind? Can everything in his head really be reduced to artificial templates and models that carry for him the burden of meaning he seeks in his films? 

Improvisation is fundamentally at odds with the logic of artificial intelligence, which produces clear texts and answers or narratives based on fixed templates and patterns. In such a system, the script is not alive, not subject to change at every moment. The script is “a living, fragile structure, always changing.” 

There is always an obscure space of doubt and reflection in the director’s consciousness that cannot be measured through technical simulation, and spirit in films cannot be generated by advanced tools. A good film, as American director Orson Welles put it, “cannot be good unless the camera is an eye in the head of a poet.” (7) The strength of a film comes only from the human sensibility of its maker. 

Today, artificial intelligence can gather thousands of filmed scenes from Gaza or Sudan, for example, sort and classify them into multiple themes, create synthetic images, and produce entire scenes that are close to reality. But can it offer the viewer a human experience that reflects the deep bond between director and film,  such as beginning with the image of a smiling cat before scenes of war? 

It is this deep bond, carrying the director’s vision, consciousness and emotional engagement with the event, that gives the film its spirit and meaning and makes it more intimate for the viewer. 

Is artificial intelligence the enemy of documentary film? 

Documentary film is built on a relationship of trust between its director and its audience. In a world increasingly surrounded by advanced tools, that trust is now being put to a major test, creating an urgent need for transparency and disclosure about the limits of AI use. 

What gives documentary its value is its objective authenticity and the truthfulness of its stories and images,  its ability to present truth faithfully, not what can be artificially manufactured. This does not mean ignoring the development of artificial intelligence. A director who fails to keep pace with technological change will be consigned to oblivion, especially since AI has opened astonishing horizons that facilitate film production: accelerating research, transcribing and analyzing texts, sorting images and improving their quality, processing sound, and reconstructing scenes from past events for which no archive exists. 

The real problem lies in the degree of reliance placed upon it: when it ceases to be an assistant and becomes a substitute mind, when the director comes to depend on it as the maker of his vision and the primary engine in constructing his films, allowing technique to overshadow substance and story. Iranian director Abbas Kiarostami once said: “I can watch films and speak about how beautiful they are technically, but I am not impressed by technique.” (8) 

A director who does not devote himself to working on his film from its inception, who does not review the footage he has shot, contemplate it, watch every image in the film’s archive, and think through the script moment by moment until the film finally sees the light, will forfeit a unique human experience capable of involving the audience in that same experience when watching the film. It is, in a sense, a renunciation of the ethical responsibility that documentary filmmaking imposes. 

References 

1. Mikhail Romm, Talks on Film Directing, translated by Adnan Madanat (Beirut: Dar Al-Farabi, 1981), p. 250.  
2. Mikhail Romm, Talks on Film Directing, translated by Adnan Madanat (Beirut: Dar Al-Farabi, 1981), p. 251.  
3. Bashar Ibrahim, “Scenes from the Occupation in Gaza 1973,” Al-Hiwar Al-Mutamaddin, Issue No. 1402, December 17, 2005.  
4. Barry Hampe, Making Documentary Films, translated by Nasser Wannous (Abu Dhabi: Department of Culture and Tourism – Dar Al-Kutub, 2011).  
5. Mikhail Romm, Talks on Film Directing, translated by Adnan Madanat (Beirut: Dar Al-Farabi, 1981), p. 252.  
6. Andrei Tarkovsky, Sculpting in Time, translated by Amin Saleh (Beirut: Arab Institution for Studies and Publishing, 2006), p. 126.  
7. Orson Welles, “Ribbon of Dreams” (originally published in 1958), republished by Sabzian, May 6, 2015.  
8. They Shoot Pictures, Don’t They — on Abbas Kiarostami.