Fascinations: AI

A briskly sarcastic essay about the layoffs of rank-and-file workers whose jobs are replaced by machines, it combines hand-crafted 16mm film with artificial intelligence, machine learning and the voices of influential figures presenting vomitous managerial monologues in verse. While the audio speeches build a dam against moral scruples, the CEOs responsible for “bad decisions” on behalf of the company are exposed and stripped by the images.“Art and commerce both require labor to make products for their respective markets. Creators do not control the markets that sell their work, which does not value labour. Experimental film is not considered collectible by the wealthy, making it unviable for the art market.”

The intricate structure of the film seems to unfurl in several tangled tentacles: one strand is made up of liminal ideas of non-human or more-than-human subjectivities, another layer is a lecture attacking institutional forms of knowledge production and preservation delivered by a man who is all too familiar with them, and the third layer is a poetic artificial intelligence generating texts based on the work of Samuel Beckett.“Some [scientists] go so far as to say that [cephalopods] have nine brains: one central one, and then all eight arms have enough neurons to allow ‘independent’ operation. My work and research with cephalopods have been mimicking this kind of decentralized ethos.”Zdroj citátu: Bomb magazine.

The personal features of the speaker, who reads a manifesto calling for the rejection of visibility, remain hidden behind an amorphous dark pink mask and an artificial voice. A montage of promotional and news spots about technologies capable of recognizing faces and storing biometric data accompanies this digital voice-over declaring disobedience. “The Queer Mask,” created from biometric data of the faces of many queer men, warns of the consequences of using these technologies as “gaydars”.“In his Manifesto Contrasexual (2001), Paul Preciado advances the queer concept of ‘contrasexuality.’ Described as a refusal of sexual norms, contrasexuality prohibits any articulation of sexuality as naturalized. Indeed, speaking the word forces one to say ‘against sexuality’—that is, against an understanding of sexuality as constituted by dominating and hegemonic power… Contrasexuality is at once a refusal, and the constitution of an alternative.”Quote source: e-flux.

The digital imprint of Karel Gott's voice was created from the sound recordings of Gott's show Zpátky si dám ten film broadcast on Czech Radio Two in 2011–2015. The obtained material was used to train a speech synthesis model based on deep neural networks. The artificial intelligence trained in this way retrieved selected passages from Karel Gott's autobiography, which were thoroughly checked by the voice synthesis editors. The Gott Forever project was co-produced by Czech Radio and Karel Gott Agency and in cooperation with SpeechTech, s.r.o. and the Research Center NTIS ZČU in Pilsen. “In the case where we are creating a digital voiceprint of a deceased person, we have set the rules very strictly. Of the basic ones, the most important is that we can only use the voice created by artificial intelligence for texts that the person actually wrote or spoke. And the other important thing is that we are not allowed to take those sentences out of context and use them for any purpose other than what they were intended for.” — Ondřej Nováček, Programme Director of Czech RadioQuote source: iROZHLAS.

The film's title refers to Harun Farocki's video installation I Thought I Was Seeing Convicts (2000), which points to the moral vacuum created around the gaze of “seeing machines”. The actors in the subversively appealing I Thought I Was Hearing Citizens are animated avatars who speak convincingly and with traces of bodily discomfort about the dangers associated with the collection and use of personal data over which we have no control.“If we automate decision-making // If we algorithmise education // If we gamify citizenship // Then who do we become?”Quote source: Manu Luksch webpage.

„The world's first short film to be 100% created, animated and narrated by artificial intelligence,“ proclaims the caption at the beginning and end of the film, which features a cast of touching, seductive and latently aggressive fictional characters generated by the Open AI application DALL-E 2 from a random set of pixel-noises. In their talks, they touch on pressing questions and anxieties about the future of the creative industries and artistic production after they are overwhelmed by the fruits of AI.“AI will not replace you; a person using AI will.” — Rodger WerkhovenQuote source: Linkedin.

The oppressive stream of seamless transformation of one photographic portrait from the artist's home archive to another – from a baby to an adult and back again somehow – shows the fragility and elusiveness of an identity that remains – or does not remain? – identical across transformations. The indistinct boundary between growth and decay is the subject of GAN (Generative Adversarial Network), which Claudia Larcher has saturated with 350 photographs of her face, accompanied by a collage of snippets from identity conversations she has had with various chatbots.“I also experiment with the audience, namely whether they perceive something as real or how far I can go until they recognize the ‘fake’.”Quote source: niio

The unapologetic monologue of a Christmas ornament who, to her own surprise, has regained consciousness, takes us through her life, from her memories of her birthplace in a Siberian holiday decoration factory, to her transport in a box and her encounter with another, to her grounding in the rhythms of a Soviet household and her unwitting emigration. A strong Russian accent lends tenderness and humanity to the robotic voice through which the shiny cosmonaut figure tentatively gropes at historical events and his own fate.“I am trying to create a futuristic poetics of data, AI and space travel. As meditations on the universe and future life on Earth, I see these as radical alternatives to the invasive agendas of private or corporate enterprises and government programmes, such as the exponential rise in human-data harvesting which are driving us towards the Singularity and societal control…”Quote source: Frieze

How fast is fast? Every 100 seconds, the perspective of the film slows down ten times, allowing us to see world-transforming events in a new perspective. Ecstatic journey that evolves every time the film is screened.

The busy shopping street Essex Road in North London, full of signs, advertisements and signage, is devoured by the confused camera “eyes” of a smartphone translator Word Lens set to convert French into English. Its vision, overwhelmed by promising stimuli, helpfully transforms the signs into more or less absurd versions of itself. A metropolitan symphony in algorithmic translation.“Thrashing around hopelessly in its dictionary, the app’s stabs in the dark replace words on shopfronts and displays with some very wayward guesswork, before it faints completely and enters They Live territory, planting spurious nouns and verbs on blank concrete walls and pavements. The empire of signs is overthrown, its language of control scrambled beyond repair…”Quote source: Tim Hayes, Sight and Sound, September 2016.

The smooth stream of transformations of the amorphous television screen, which grows with the frame into a new, even larger television set, which again becomes the content of the medium, underlines the image of television as an all-consuming organism that eternally reproduces itself. The hypnotic homage to the media's endless self-reflection is animated by artificial intelligence.“I have been obsessed with programming since it was introduced to me in the midst of the 90s Finnish demoscene, a UNESCO immaterial world heritage-listed computer subculture. In 2009, I opened a portal to other dimensions in every given JPEG by inventing a pixel-sorting style algorithm as my MFA graduation work. Ever since then, I have been studying variations of it.”Quote source: Kultama.

A member of the art collective Philth Haus SYLLA is an artificial intelligence creating videos based on the mass media representation of “girlhood”. A poem about the desire for self-determination and acceptance, which is heard several times in the film, is accompanied by advertising stills that remain literal and illustrative, although they gradually lose their “authentic” humanity. SYLLA is largely inspired by the work of the French philosophical collective Tiqqun, the book Preliminary Materials for a Theory of the Young-Girl (1999), which discusses the girl as a “total product of consumer society”.“The Young-Girl is not always young; more and more frequently, she is not even female. She is the figure of total integration in a disintegrating social totality.”Quote source: Theory of the Young-Girl by Tiqqun.

The image of a shabby façade, whose transformations are set in motion by artificial intelligence, loses one clue after another and ruthlessly tests our ability to recognise shape. The film was co-created with Claudia Larcher by Generator and Discriminator, neural networks that fed off the paper collages Baumeister (2011–2021), which Larcher created from clippings from architectural magazines. A soundtrack of unstable rippling images was derived by “helpers” called Dynascore and Imaginary Soundscape.“I think the screen is a heterotopia in and of itself, a space that becomes “other” and inserts itself into our living environment.”Quote source: niio