Telepathic Letters. Courtesy of the filmmakers
In recent decades, the Locarno Film Festival has established itself as a premiere market for some of the more unusual experiments to come through the festival circuit. In some of the films, the image is composed of a heterogeneous mix of various formats and media; in others, the textures of the visual surface, from digital artifacts to film grain, supersede the content within. In a conversation with video essayist and film scholar Kevin B. Lee in 2023, Lee identified this proliferating tendency as a possible side effect of social media, which consists of sounds and images that are “less dependent on narrative,” the effects of which on our visual literacy “may have influenced cinema’s development into something more about textures, moments, and vibes.”
Within the context of nonfiction cinema, specifically, this focus on low-resolution images, digital artifacts, and film grain can be seen as a reaction to the rise of prestige documentaries on streaming platforms like Netflix and HBO during the past decade, the central tenets of which are uniformly high digital production values and heightened narrativization. Netflix’s (in)famous list of approved cameras, for instance, was justified by the company in the following manner: “One of the biggest priorities for us as a studio is helping our filmmakers do their very best work. We want our filmmakers to not just feel enabled but also encouraged to use the latest and greatest capture technologies out there to tell their stories.” One needs not so much to read between the lines than read them as they are to notice the emphasis on storytelling and a uniform, high-fidelity audiovisual presentation. In Netflix's case, the most expensive subscription tiers have perks such as 4K and HDR. The means for platforms to incentivize patrons to upgrade are found in the above-cited “encouragement” of the filmmakers to use “the latest and greatest capture technologies to tell their stories.” Seemingly neutral economic considerations exert not only ideological but also powerful aesthetic effects on resultant works.
For the films at Locarno, what’s at stake is not what we see, nor the act of seeing itself, but the instruments that mediate that act: the technologies of vision. The opening title of Telepathic Letters identifies it as “v. 23 (Locarno mix)” and “an unfinishable film by Edgar Pêra.” It’s a speculative conversation between deceased writers H. P. Lovecraft and Fernando Pessoa based on passages taken from each author’s oeuvre, repurposed and organized by the film into an imaginary correspondence. Pêra sets himself a supernatural challenge, as both subjects of his epistolary film have been dead for decades. To resurrect them, he enlists the services of text-to-video generative Artificial Intelligence (genAI), which is responsible for all of the visuals in Telepathic Letters. As we hear voice actors recite lines from the two writers, the visuals alternate between the genAI visions of what’s being read (e.g., Lovecraft’s fantastical monsters, hallucinatory corridors of abstract shapes, eerie dystopian landscapes, or suited marching bears who are playing the drums ) and similarly generated visualizations of the authors speaking the lines. Both are often represented in split-screen panels in various iterations, corresponding to Pessoa’s concept of “fragmented identities” (individuals can have multiple personas) and Lovecraft’s numerous pseudonyms and alternate personas. The objective of Telepathic Letters unfolds in finding parallel ideas between the writings of two seemingly unrelated authors. As the narrator, a female voice with its own AI-generated visualization, states, “We will uncover the hidden connections between their unique points of view.”
But how does AI imagine the cryptic lines of these writers? André Bazin saw cinema’s distinct ontological faculty in the camera’s capacity for autonomous and automatic reproduction of objective reality. Emerging as a new tool for filmmaking, with inherent qualities distinct from that of the camera, genAI alters the relationship between reality and representation, between the filmmaker and the world, and between the subjective and the objective. In other words, it complicates and alters our understanding of cinematic ontology. In the film, Lovecraft says, “The universe may be a dream, but it cannot be considered a human dream.” Similarly, here we are faced with the seemingly automatic production of a subjectivity that neither belongs to any person in particular nor is immediately dependent on objective reality. To put it in the language of Pessoa and Lovecraft, AI produces a collective dream, a concretized amalgamation of hitherto existing visualizations.
While intriguing parallels between the ideas of Pessoa and Lovecraft are teased throughout the film, Letters’ use of the above-mentioned tension primarily operates on a sensory plane and remains on the surface. A quote from Pessoa may best sum up the film’s attitude: “There’s nothing, no reality, but sensations. Ideas are sensations, but of things not placed in space and sometimes not even in time. Dreams are sensations with only two dimensions. Ideas are sensations with only one dimension. We must create an art with only one dimension.” Letters takes this to heart, not just in constructing a hallucinatory experience but in creating one that’s fairly one-dimensional. Despite the complex aesthetic, theoretical, economic, and moral implications of AI-generated images for cinema, the approach and results in Letters are straightforward, with the visuals simply illustrating the narration in the way that archival images are used to demonstrate a fact in a traditional documentary. The result, with the visualizations of the narrator, Pessoa, and Lovecraft, is comparable to short AI-generated videos you may find on TikTok or YouTube, with the characters captured through a (virtual) wide lens, more or less centered in the frame, looking either at or slightly off-camera. A similarly uncomplicated approach can be seen in the uncanny sights of transmuting figures and environments, the dark color palette, the oppressive atmosphere, and the unnerving music and sound design, all constants throughout the film. Novel tools like AI can open up new ontological possibilities for cinema. But they will remain mere possibilities so long as they are used to emulate aesthetics and structures that grew out of the particular needs of lens-based cinema.
If the central question evoked by Letters is the extent to which a parallel could be drawn between the unconscious imagination of humans (dreams) and technology (Artificial Intelligence), Adele Tulli’s Real examines the capacity of consciously imagined virtual worlds to replace the real world. Or, as a character in the film, whom we only see represented by their VRChat avatar, says, “Not the real world, but the physical world. This world feels very, very real to me.” Real weaves together footage from a panoply of image-producing/capturing/streaming technologies: VRChat, Google Earth, GoPro, YouTube, video games, Zoom, OnlyFans, and surveillance cameras. Throughout the film, we observe slices from the lives of a few individuals interconnected with technology: a young Korean boy asks Bixby (Samsung’s virtual assistant) if the AI thinks he’s cute. A trans couple who’ve never seen each other physically find affirmation of their identities in VRChat, which they describe as “a utopia beyond the bounds of capitalism.” An array of influencers and YouTubers come to terms with their mental health problems and admit to suffering from depression underneath the cheerful facade they need to put up. The passengers of a Venetian gondola are all bewitched by their phones amidst centuries of art and culture. A delivery driver who rides around with a live video stream says, “If you do delivery at night, you make a lot of money, but you feel lonely.”
The film’s message is clear enough: we can’t seem to connect with the world without the mediation of technology. Amongst the proliferation of technologies that produce and circulate images, humans are increasingly experiencing atomization and loneliness. However, Real suggests that while technology can isolate us, it can also connect those in isolation. This is a correct, if fairly self-evident assertion. But in addition to the more explicit theme of reality’s seemingly inescapable mediation through technology, another, more intriguing theme is teased: that of the image no longer as a photograph but as a multiplicity of heterogeneous surfaces that mediate reality through transforming, representing, reflecting, or capturing it. The film explicitly draws attention to this, as shown by a sequence in which an orgy of various images (phones being held by disembodied hands, influencers with effects on their faces, groups performing trendy dances, etc.) march in a virtual tunnel. However, except for this superficial attempt, it doesn’t develop this idea in a meaningful direction; instead, it focuses on the more palpable (if also more mundane) theme of mediating reality discussed before.
One thing the film avoids in its investigation of the technological mediation of reality is the unacknowledged, naturalized camera of the filmmakers. Real’s scenes tend to be composed of shots from webcams, security cameras, Google Earth, etc., seamlessly transitioning to shots from the filmmakers’ camera, which then carries the scene forward. In this act, Real differentiates between all the other image-capturing technologies (its subjects of investigation) and the camera(s) used to capture them (its object of investigation), which is treated as neutral and natural. The presence and effect of the filmmaker’s camera has been a long-debated issue in discussions surrounding documentary films. One must ask whether the proliferation of other imaging technologies, less “natural” and familiar than the camera, could encourage obliviousness to cinema’s predominant technology of vision.
Arta Barzanji is a London-based Iranian filmmaker, critic, and lecturer. His current film project is the documentary Unfinished: Kamran Shirdel. Arta is an alumnus of the critic programmes at the Locarno and Ghent film festivals and has written for outlets including MUBI Notebook, Sabzian, and photogėnie.