DEEP EATING
(2017) INSTALLATION (PHOTOGRAPHY)
In 1990, John Divola, in his "The Green of this Notebook", used existential poetics to address the essential nature of the photographic process of a pre-digital era. The series of images corresponded to the reference on pages from Sartre's "Being and Nothingness" with some visual narrative based on an appreciation of the subjective and expressive power of the examples from reality.
'Deep Eating' focuses on radically different actors: a machine in the object recognition deep learning process and the reference to the resulting judgment of the existing data provided by the Google Books library.
If machine sense does exist, it exists in useless texts. Encoded with irrelevant illustrations. Divola visualised text which implemented the Western cultural code. Playing the other way around, by verifying a new value of a random choice from the speech abyss, one can believe in penetrating another modality of seeing.
If machine sense does exist, it exists in useless texts. Encoded with irrelevant illustrations. Divola visualised text which implemented the Western cultural code. Playing the other way around, by verifying a new value of a random choice from the speech abyss, one can believe in penetrating another modality of seeing.
In the late 50s (the time of active struggle against racial segregation in The United States), French philosopher Gilbert Simondon wrote his paper "On the Mode of Existence of Technical Objects", where he argues that culture must come to terms with technical entities as part of its body of knowledge and values. He proposes the recognition of the modes of existence of technical objects to be the
result of philosophic consideration and compares this process with analogous to what the abolition of slavery achieved in affirming the worth of the individual human being ( Simondon 1958).
In the process of overcoming phenomenological subjectivity, most scholars were focusing on revising an animal's position as a speechless and subaltern entity, which doesn't even experience pain (according to Descartes). Inevitably to reinvent the understanding of a human, the main human features started to be rifled through. Thereby, Georgio Agamben named language as such distinguisher, the phenomenon not given naturally or inherent in the psychophysical structure of man but produced historically with no proper assignment neither to man nor to an animal. "If this element is taken away, the difference between man and animal vanishes, unless we imagine a nonspeaking man - Homo alalus, precisely— who would function as a bridge that passes from the animal to the human" ( Agamben 2002). This Homo alalus - the figure of exclusion - means a perfect example of bare life: a deprived state of non-achievement of human's (according to Heidegger) Dasein stage. Derrida, distinguishing the absolute other under the gaze of his cat, pondering about its ability to respond: "When it responds in its name (whatever respond means and that will be our question), it doesn't do so as the exemplar of a species called cat, even less so of an animal genus or realm" ( Derrida, Wills 2002). He's focusing on the name given to a cat, but before the name, he recognizes an animal itself, whom Derrida is following after with the notion of his nakedness. Anyway, the question of speech remained open, but the vague hints of overcoming subjectivity are fumbled: to be is to be with, after, alongside, before: in different modes of being but with an animal. Uexküll, much focusing on a temporal and morphological understanding of animals, was not keen to compare them to machines: an animal was, at best, an imperfect machine (Parikka 2009). Nevertheless, his proposed distinction of "the Umgebung, the objective space in which we see a living being moving, from the Umwelt, the environment-world that is constituted by a more or less broad series of elements that he calls "carriers of significance" (Bedeutungsträger) or of "marks" (Merkmalträger), which are the only things that interest the animal" (Agamben 2004) is desirable to imply on machines, since they achieved proper complexity, so the discussion about Artificial Intelligence became possible. The uniqueness of a relation man-machine is accumulated in the fact that, following Simondon, "the machine with a superior technicality is an open machine and the ensemble of open machines assumes a man as a permanent organizer and as a living interpreter of the interrelationships of machines." He comes up with a beautiful comparison of a man with a conductor: "far from being the supervisor of a squad of slaves, man is the permanent organizer of a society of technical objects which need him as much as musicians in an orchestra need a conductor. The conductor can direct his musicians only because, like them and with a similar intensity, he can interpret the piece of music performed; he determines the tempo of their performance, but as he does so, his interpretative decisions are affected by the actual performance of the musicians; in fact, it is through him that the members of the orchestra affect each other's interpretation; for each of them he is the real, inspiring form of the group's existence as a group; he is the central focus of interpretation of all of them in relation to each other. This is how a man functions as a permanent inventor and coordinator of the machines around him. He is among the machines that work with him". This openness being noticed by Uexküll in animals became a key factor of Deleuze and Guattari's texts, where they speak of positing a plane of immanence on which the issue of categorical differences between animals and humans, nature and technology is bracketed and the view of affects, movements, and relations among parts is posited as primary (Parikka 2010). Furthermore, Deleuze extends this plane of immanence to a technique of nature,
in which "artifice is fully a part of Nature, since each thing, on the immanent plane of nature, is defined by the arrangements of motions and affects into which it enters, whether these arrangements are artificial or natural" (Deleuze, Guattari 2015). This statement justifies the supposed chance to enter the plane of immanence through an interaction with a machine that has to stop, recalling Simondone being an unrecognized, materialized, and enslaved stranger. Since personalities and entities are studied in motion, in the transition, a machine is something that slows down entropy and stabilizes the world.
result of philosophic consideration and compares this process with analogous to what the abolition of slavery achieved in affirming the worth of the individual human being ( Simondon 1958).
In the process of overcoming phenomenological subjectivity, most scholars were focusing on revising an animal's position as a speechless and subaltern entity, which doesn't even experience pain (according to Descartes). Inevitably to reinvent the understanding of a human, the main human features started to be rifled through. Thereby, Georgio Agamben named language as such distinguisher, the phenomenon not given naturally or inherent in the psychophysical structure of man but produced historically with no proper assignment neither to man nor to an animal. "If this element is taken away, the difference between man and animal vanishes, unless we imagine a nonspeaking man - Homo alalus, precisely— who would function as a bridge that passes from the animal to the human" ( Agamben 2002). This Homo alalus - the figure of exclusion - means a perfect example of bare life: a deprived state of non-achievement of human's (according to Heidegger) Dasein stage. Derrida, distinguishing the absolute other under the gaze of his cat, pondering about its ability to respond: "When it responds in its name (whatever respond means and that will be our question), it doesn't do so as the exemplar of a species called cat, even less so of an animal genus or realm" ( Derrida, Wills 2002). He's focusing on the name given to a cat, but before the name, he recognizes an animal itself, whom Derrida is following after with the notion of his nakedness. Anyway, the question of speech remained open, but the vague hints of overcoming subjectivity are fumbled: to be is to be with, after, alongside, before: in different modes of being but with an animal. Uexküll, much focusing on a temporal and morphological understanding of animals, was not keen to compare them to machines: an animal was, at best, an imperfect machine (Parikka 2009). Nevertheless, his proposed distinction of "the Umgebung, the objective space in which we see a living being moving, from the Umwelt, the environment-world that is constituted by a more or less broad series of elements that he calls "carriers of significance" (Bedeutungsträger) or of "marks" (Merkmalträger), which are the only things that interest the animal" (Agamben 2004) is desirable to imply on machines, since they achieved proper complexity, so the discussion about Artificial Intelligence became possible. The uniqueness of a relation man-machine is accumulated in the fact that, following Simondon, "the machine with a superior technicality is an open machine and the ensemble of open machines assumes a man as a permanent organizer and as a living interpreter of the interrelationships of machines." He comes up with a beautiful comparison of a man with a conductor: "far from being the supervisor of a squad of slaves, man is the permanent organizer of a society of technical objects which need him as much as musicians in an orchestra need a conductor. The conductor can direct his musicians only because, like them and with a similar intensity, he can interpret the piece of music performed; he determines the tempo of their performance, but as he does so, his interpretative decisions are affected by the actual performance of the musicians; in fact, it is through him that the members of the orchestra affect each other's interpretation; for each of them he is the real, inspiring form of the group's existence as a group; he is the central focus of interpretation of all of them in relation to each other. This is how a man functions as a permanent inventor and coordinator of the machines around him. He is among the machines that work with him". This openness being noticed by Uexküll in animals became a key factor of Deleuze and Guattari's texts, where they speak of positing a plane of immanence on which the issue of categorical differences between animals and humans, nature and technology is bracketed and the view of affects, movements, and relations among parts is posited as primary (Parikka 2010). Furthermore, Deleuze extends this plane of immanence to a technique of nature,
in which "artifice is fully a part of Nature, since each thing, on the immanent plane of nature, is defined by the arrangements of motions and affects into which it enters, whether these arrangements are artificial or natural" (Deleuze, Guattari 2015). This statement justifies the supposed chance to enter the plane of immanence through an interaction with a machine that has to stop, recalling Simondone being an unrecognized, materialized, and enslaved stranger. Since personalities and entities are studied in motion, in the transition, a machine is something that slows down entropy and stabilizes the world.
A little while ago, I started a project that I can assign as performative or collaborative. It involves my part as a creator of the initial image – a straight, nonmodified, digital, and partly abstract photograph; then, I suggest a machine interpret my proposal according to its own decision. The result, which appears in a text form, then helps to find a match throughout a multiplicity of text sources. The "Deep Eating" is a visual homage to John Divola's "The Green of this Notebook". This work was made in the late 1990s on the West Coast of the United States and closely relates to individual artistic experiences generalized further. Divola says his project uses existential poetics to address the essential nature of the photographic process. It is based on an appreciation of the subjective and expressive power of the examples themselves and the specifics of Sartre's context. In "The Green of this Notebook", Divola presents a series of images corresponding to the reference on pages from Sartre's "Being and Nothingness" with some visual narrative. The book opens with the highlighted text, "I am on a narrow path – without a guard rail – which goes along a precipice." The literal relationship of image/text combinations continues throughout the book, ending with "If I eat pink cake, the taste of it is pink." According to Divola, this particular text is selected because it is an extreme and ambitious case of moving from the specific to the general and because of the richness of Sartre's supporting examples.
It's not much to say in this case that Sartre's "Being and Nothingness" (1943), A Phenomenological Essay on Ontology, is a reply to Heidegger's ontology he formulated in "Being and Time" which formulates the Subject-oriented notion about existence preceding essence. Divola meditates on this notion through the idea of the index nature of photography, hesitating from the concrete image to a general utterance.
In 2015, Microsoft Corporation published a report on an app based on several neural nets and using deep learning technology. Its name is Microsoft Caption Bot, and it appears to be an image caption system that automatically describes images in the wild. The core idea of the app is the inclusion of high-quality caption quality with respect to human judgments. The report claims that image captioning is a fundamental task in Artificial Intelligence that describes objects, attributes, and relationships in an image in a natural language form. It has many applications such as semantic image search, bringing visual intelligence to chatbots, or helping visually-impaired people to see the world around them. Caption Bot consists of a pair of neural networks, one for mapping each input modality (verbal and visual) to a common semantic space. They are trained jointly to make sets of image/caption pairs. The app partly exists as an open-source where any Internet user can contribute in detecting and analyzing the result of computer captioning by ranking from 1 to 5. The app describes itself as a young entity on a stage of learning elaborate pattern recognition. When using this app, one could notice some recognition mistakes occurring. The first association which appears is a comparison to grammar being learned by a child or a foreigner. They tend to grasp the main language algorithm and roughly apply it to every situation, ignoring nuances and less notable and exquisite rules. Yet, this mistake overview tells much about the language itself, where it stems from, and how it develops.
From the human perspective, one can say that the above-mentioned machine is not in the stage of grammar or semantic cognition. Noam Chomsky revealed this crucial difference, who researched grammar basic carcasses and invented an example: "Colorless green ideas sleep furiously" (1957), grammatically correct but semantically nonsensical. Not superfluous will remind us of the semantic meaning of the images proposed. It has already been mentioned that machines are learning toward human judgment. This judgment is designed according to a hierarchy of human needs (face detecting, hazard recognition, alert registration). This means the human language is definitely based on a human logic of well-being/ survival. But it can mismatch the machine logic operating another kind of algorithm. What can we say about language attributed to a machine in this case? Or even language as merely a human invention which represents a symptom (according to Lacan), thus can be only conditioned on a correlation of biological, more precisely, higher nervous activity premises? The words coming after image analysis resemble human language but can appear not what they seem. Either this process expands Derrida's concept of re-mark, or it takes this language ready-made and uses it as a superstructure impending over a code: as an interface to be more or less understandable by a human.
Can we claim that the machine has its own Umwelt? To reveal this, at least to try to enter it, I supply a machine with different images constructed depicting eatable elements. I eliminate the normal appearance of objects and deconstruct the image to the threshold of abstraction. Nonetheless, the human's eye remains still capable of detecting the deconstructed image, it focuses on the materiality of an organic substance, an eatable substance faster than on the surrounding elements. The machine has another logic. Sometimes it's even imperceptible to explain it, but, for sure, it doesn't come out randomly. If we reverse the sequence of conversion, it will give an awareness of how differently the machine can perceive the world around it. Even if it starts to return a "gaze" to yourself, not necessary it will "see" a human before it. A person who suffers from schizophrenia (which is not accidentally called a disease of language) can take people around for non-living objects. The best way to prove it is to give a machine to illustrate some insignificant texts, unverified plots, and unknown books – an abyss of rather useless texts one can find in the flow of, for instance, Google Books (As of October 2015, the number of scanned book titles was over 25 million, Google estimated in 2010 that there were about 130 million distinct titles in the world, and stated that it intended to scan all of them). If machine sense does exist, it exists in useless texts. Encoded with irrelevant illustrations. John Divola visualized text which implemented the Western cultural code. Playing the other way around, by verifying a new value of a random choice from the speech abyss, one can believe in penetrating another modality of seeing.
It's not much to say in this case that Sartre's "Being and Nothingness" (1943), A Phenomenological Essay on Ontology, is a reply to Heidegger's ontology he formulated in "Being and Time" which formulates the Subject-oriented notion about existence preceding essence. Divola meditates on this notion through the idea of the index nature of photography, hesitating from the concrete image to a general utterance.
In 2015, Microsoft Corporation published a report on an app based on several neural nets and using deep learning technology. Its name is Microsoft Caption Bot, and it appears to be an image caption system that automatically describes images in the wild. The core idea of the app is the inclusion of high-quality caption quality with respect to human judgments. The report claims that image captioning is a fundamental task in Artificial Intelligence that describes objects, attributes, and relationships in an image in a natural language form. It has many applications such as semantic image search, bringing visual intelligence to chatbots, or helping visually-impaired people to see the world around them. Caption Bot consists of a pair of neural networks, one for mapping each input modality (verbal and visual) to a common semantic space. They are trained jointly to make sets of image/caption pairs. The app partly exists as an open-source where any Internet user can contribute in detecting and analyzing the result of computer captioning by ranking from 1 to 5. The app describes itself as a young entity on a stage of learning elaborate pattern recognition. When using this app, one could notice some recognition mistakes occurring. The first association which appears is a comparison to grammar being learned by a child or a foreigner. They tend to grasp the main language algorithm and roughly apply it to every situation, ignoring nuances and less notable and exquisite rules. Yet, this mistake overview tells much about the language itself, where it stems from, and how it develops.
From the human perspective, one can say that the above-mentioned machine is not in the stage of grammar or semantic cognition. Noam Chomsky revealed this crucial difference, who researched grammar basic carcasses and invented an example: "Colorless green ideas sleep furiously" (1957), grammatically correct but semantically nonsensical. Not superfluous will remind us of the semantic meaning of the images proposed. It has already been mentioned that machines are learning toward human judgment. This judgment is designed according to a hierarchy of human needs (face detecting, hazard recognition, alert registration). This means the human language is definitely based on a human logic of well-being/ survival. But it can mismatch the machine logic operating another kind of algorithm. What can we say about language attributed to a machine in this case? Or even language as merely a human invention which represents a symptom (according to Lacan), thus can be only conditioned on a correlation of biological, more precisely, higher nervous activity premises? The words coming after image analysis resemble human language but can appear not what they seem. Either this process expands Derrida's concept of re-mark, or it takes this language ready-made and uses it as a superstructure impending over a code: as an interface to be more or less understandable by a human.
Can we claim that the machine has its own Umwelt? To reveal this, at least to try to enter it, I supply a machine with different images constructed depicting eatable elements. I eliminate the normal appearance of objects and deconstruct the image to the threshold of abstraction. Nonetheless, the human's eye remains still capable of detecting the deconstructed image, it focuses on the materiality of an organic substance, an eatable substance faster than on the surrounding elements. The machine has another logic. Sometimes it's even imperceptible to explain it, but, for sure, it doesn't come out randomly. If we reverse the sequence of conversion, it will give an awareness of how differently the machine can perceive the world around it. Even if it starts to return a "gaze" to yourself, not necessary it will "see" a human before it. A person who suffers from schizophrenia (which is not accidentally called a disease of language) can take people around for non-living objects. The best way to prove it is to give a machine to illustrate some insignificant texts, unverified plots, and unknown books – an abyss of rather useless texts one can find in the flow of, for instance, Google Books (As of October 2015, the number of scanned book titles was over 25 million, Google estimated in 2010 that there were about 130 million distinct titles in the world, and stated that it intended to scan all of them). If machine sense does exist, it exists in useless texts. Encoded with irrelevant illustrations. John Divola visualized text which implemented the Western cultural code. Playing the other way around, by verifying a new value of a random choice from the speech abyss, one can believe in penetrating another modality of seeing.