HOW TECHNOLÓGOS “RESPONDS” TO WHAT USED TO BE CALLED “IMAGES.” A MEDIA-ARCHAEOLOGICAL RESPONSE TO THE “QUESTIONNAIRE ON THE CHANGING ONTOLOGY OF THE IMAGE”

LIBERATING THE IMAGE FROM ITS ANTHROPOCENTRIC DEFINITION “Traditionally we think of images as [...] delimited phenomena that in one way or the other appear to the human mind and apparatus of perception” (Questionnaire). The choice of words in the Questionnaire is indicative already. When optical physiology and cognitive image sensation—from the “analogue” camera obscura-like eye to the almost “digital” signal-computing brain— is observed closely,1 image processing within the human turns out as, indeed, a function of an “apparatus.” Sigmund Freud’s nonmetaphorical concept of the psychic “Apparat” in chapter VII of his Interpretation of Dreams2 explicitly compares the preliminary stages of imaging to the microscope, or to photography.3 The mechanistic approach reemerged in protocybernetic research into the electrical circuit simulation of neural image perception.4 The human “mind and apparatus of perception” (Questionnaire) literally became a nonhuman machinery in Rosenblatt’s computational Perceptron, liberating the “image” from its physiological anthropocentrism.5 Machine vision, so far, stayed profoundly different from human image cognition. But technical images as outputs from Artificial Neuronal Nets start to challenge, and to emulate, the human imaginative potential, once they are not only trained by human tagging, but (in a more complex way) by rivalling machines among themselves which are fed with big data derived from “social media.” Just like Gottfried Ephraim Lessing, in his 1766 treatise Laokoon, had almost identified the aesthetic properties of the visual arts as parallel perception (aisthesis, in the Aristotelean sense) of coexistent units in space, today, it is no coincidence that “deep” machine learning takes place in parallel graphics processing units (GPUs) that were originally developed for image processing in computers. Artificial Intelligence does not simply mimick human image perception (even if Van Gogh-like paintings HOW TECHNOLÓGOS “RESPONDS” TO WHAT USED TO BE CALLED “IMAGES.” A MEDIA-ARCHAEOLOGICAL RESPONSE TO THE “QUESTIONNAIRE ON THE CHANGING ONTOLOGY OF THE IMAGE”

and cognitive image sensation-from the "analogue" camera obscura-like eye to the almost "digital" signal-computing brainis observed closely, 1 image processing within the human turns out as, indeed, a function of an "apparatus." Sigmund Freud's nonmetaphorical concept of the psychic "Apparat" in chapter VII of his Interpretation of Dreams 2 explicitly compares the preliminary stages of imaging to the microscope, or to photography. 3 The mechanistic approach reemerged in protocybernetic research into the electrical circuit simulation of neural image perception. 4 The human "mind and apparatus of perception" (Questionnaire) literally became a nonhuman machinery in Rosenblatt's computational Perceptron, liberating the "image" from its physiological anthropocentrism. 5 Machine vision, so far, stayed profoundly different from human image cognition. But technical images as outputs from Artificial Neuronal Nets start to challenge, and to emulate, the human imaginative potential, once they are not only trained by human tagging, but (in a more complex way) by rivalling machines among themselves which are fed with big data derived from "social media." Just like Gottfried Ephraim Lessing, in his 1766 treatise Laokoon, had almost identified the aesthetic properties of the visual arts as parallel perception (aisthesis, in the Aristotelean sense) of coexistent units in space, today, it is no coincidence that "deep" machine learning takes place in parallel graphics processing units (GPUs) that were originally developed for image processing in computers. Artificial Intelligence does not simply mimick human image perception (even if Van Gogh-like paintings HOW TECHNOLÓGOS "RESPONDS" TO WHAT USED TO BE CALLED "IMAGES." A MEDIA-ARCHAEOLOGICAL  RESPONSE TO THE "QUESTIONNAIRE ON THE CHANGING  ONTOLOGY OF THE IMAGE" Wolfgang Ernst The Nordic Journal of Aesthetics, No. 61-62 (2021), pp. 84-93 produced by AI are gimmicks, according to contemporary discourse), but more profoundly, it challenges the narcissistic view that humans are the only beings able to develop a semantic sense of imagery. The outputs of Artificial Neural Nets (such as the Self-Organizing Maps), which are predominantly images, remind the human of the technicity of his, or her, image perception itself.

THE "NETWORKED IMAGE": TEXTUALITY, LITERALLY
The current "intensification of what we might call the networkedness of the image" (Questionnaire), from a media-archaeological point of view, is an externalization of the inherently "textile" essence of technical image production. 6 The technical image within media-archaeologies has already been networked from within, both materially and logically: as a tissue, which became techno-logical in Jacquard's programmable loom. The general theme of the Dutch Electronic Art Festival in Rotterdam in February-March 2003, "Data Knitting," 7 reminded of this first proto-digital "image" production in France around 1800, which soon afterwards provided the model for data processing in Charles Babbage's design of a first programmable computer. It is not the high resolution of an image that is crucial for its digitized reproduction (or rather: transformation), but its addressability at every discrete pixel element. This is a nonsocial approach, since it ignores the discursive implications of, for example, the painter's intention. With digitization, what has belonged to the Humanities so far, becomes algorithmically "inhuman." Electronic face recognition identifies schemes, not How Technológos "Responds" to what used to be called "Images" individuals. Once translated into computable numbers, the memory of art from the cultural past invites for algorithmic analysis, such as pattern recognition (aka "style"), in amounts previously unattainable for a single scholar.

MACHINE VISION, AND ITS COLD (MEDIA-) ARCHAEOLOGICAL GAZE
In its escalation from passive "analogue" telecommunication to algorithmicized "digital" intelligence (in all senses), the age of electronic signal transmission, storage, and processing, has resulted in "a proliferation of machine imagery that operates independently of human perception and cognition" (Questionnaire) indeed. What cultural aesthetics used to call "image" so far, radically transforms into a techno-mathematical image function. Such technically operative (rather than bodily performative) images have become "images without a social goal" (Questionnaire). They relate to the "cold media-archaeological gaze" instead.
The cold camera-eye gaze of televisonary media relates to media-archaeological aesthetics. Friedrich Nietzsche's "pathetic distance" in philosophical analysis insists on the exteriority of analysis, as opposed to hermeneutic empathy. Even more rigorously, this corresponds with Ernst Jünger's aesthetics of detachment as a mode of perception created by optical technology. 10 The image searching software of the company Cobion in Kassel, Germany, for example, once crawled the Web for pornographic child abuse images-this task could have been painful for humans, but not for the machine. 11 The mediaarchaeological gaze is cold in McLuhan's sense of differentiating between "hot" and "cold" media-with the latter ones inviting a human receiver to participate actively when putting visual signal streams into relation(s). When media themselves become active archaeologists of data, the cold gaze of the machine is no longer an empathetic vision but an optical element in cybernetic feedback systems.
Does it make sense at all for media theory to metonymically apply the category of the human gaze to machine vision? Dziga Vertov, in his film The Man with the Camera, makes the cameraeye (the KinoGlaz) an agent of vision. In Alfred Hitchcock's film The Birds, at one point, the camera switches to the birds-eye perspective from above, making the whole scene look completely different. But the overall perspective in this film is, technically, that of the camera "eye" which is still an analogy to human panoptic perception. In techno-mathematical systems, however, the notion of seeing itself becomes metaphorical. In their processing of optical inputs as data, signals are radically abstracted. In such Media archaeology looks at digital images not iconologically, but technologically. The "cold archaeological gaze" recognizes the digital image format as border-defined functions of data manipulation, once they have been sampled into the digital regime by analogue-to-digital conversion, 14 or when they are born algorithmically. Images thus become calculable, rather than narratable.
Media archaeology addresses the technical aspects of the image, that is: "as medium." But even the term "technical image" is actually misleading, since technology does not know images in the human sense. Let the image be rather defined mathematically, as "a real-valued function of two real variables." 15 Does the "image" therefore make sense (in its double meaning) only for human phenomenology, or is there something like an "alien" image phenomenon from the technical point of view? 16 "The term 'picture' suggests a flat object whose appearance varies from point to point." 17 This variation has been a perceptual function of human vision so far, but now returns from within technology, where a spatial distribution becomes radically temporalized. 18 Already with electronic television, it has been the human eye only which finally integrates the "flying spot" emanating from the cathode rays tube into an "image"-while "[a] machine can capture the same image, without any consciousness or experience of the visual form." 19 Is there something like the technologically "implicit image"?
While human-made, body-linked images, as cultural techniques, cannot be "liberated" from cultural semantics and (art) historical iconology, 20 genuinely technically coded image actions "do not represent an object" any more (Questionnaire), "but rather are part of an operation" (Questionnaire). Such imaging is no longer primarily cultural iconology, but a truly technical iconology coming to its own, a "log-icon," 21 in Charles S. Peirce's sense of diagrammatic iconicity. is simply some in-camera image processing" which enlarges the image area at the center of the frame and trims away the outside edges. 23 The camera itself has no sense of the "image" but only knows techno-mathematical image functions. The digital "image" technologizes the visual image. Shouldn't the "visual" here be put in quotation marks, or rather the "image" itself?
In media-archaeological analysis, the former "social field" becomes a data matrix. A UK company Vivacity Labs company has installed a thousand of surveillance cameras across the country for the permanent registering of pedestrians, bikes, and cars. Once transcribed into data, such optical signals can be interpreted by an Artificial Intelligence algorithm, in order to predict and anticipate, for example, a traffic jam. 24 All of a sudden, the Corona pandemic altered the function of this panoptical dispositive of dataveillance in favour of automatically controlling "social distancing," by modifying the algorithm to measure the prescribed distance between passers-by. Thereby vivacity itself (literally) becomes a technical function-in direct analogy to the Corona-warning (or -tracing) apps in smartphones on the basis of Bluetooth distance metrics that does not capture images of humans any more, but simply the data of their mobile communication devices. Ironically, it is now the face masks (meant to hinder viral contagion) that hinder automated face recognition which has been the concern of privacy protection so far. It is only anonymous data which are thereby generated.
HUMAN IMAGE AESTHETICS VS.

ICONO-LOGICAL IMAGE PROCESSING
A truly techno-logical ontology of the image takes place in computer graphics. The most radical media-archaeological analysis is enacted by the machine itself, as an approach "that challenge[s] an anthropomorphic register" (Questionnaire).
Most images in social media are still "indexically derived" (Questionnaire) and thereby referenced to the human lifeworld.

Such digital imagery is nothing but a mere "extension of man" in
McLuhan's sense. The digital image comes to its own, and starts to develop its own ontology, only when it is "born" algorithmically.

DISAPPEARANCE OF THE "IMAGE OF MAN"
Let us finally return to the initial concern of the "Questionnaire." Not only the exclusiveness of human "image" perception is displaced by the operative image, but the image of man as the author of images itself vanishes. In the notorious final remark of his Order of Things, Foucault predicts that the image of the human will disappear like the shape of a human face drawn into the sand at the sea shore. 33 In times of highly integrated microcomputer chips, this sand is indeed silicon, which dissolves the image of man by "calculation"-literally "counting with calculi," that is: with pebbles. The corporal image of man reappears on computer screens, but temporarily disappears in the actual medium channel.
Foucault's apocalyptic metaphor becomes micro-temporalized by actual computing, which is recording, transduction, processing, transmission and storage of man as "data face." Such a machine imaging escapes human "vision" all together. All that remains, for humans, is to learn circuit diagrams, and to read code.

WOLFGANG ERNST is Full Professor for Media Theories at the Institute for Musicology and Media
Science at Humboldt University in Berlin. His current research covers "radical" media archaeology as method, the theory of technical storage, the technologies of cultural transmission, micro-temporal media aesthetics and their chronopoetic potentials, and sound analytics ("sonicity") from a media-