Algorithmic Intimacy: the Digital Revolution in Personal Relationships, Anthony Elliott (2022) Polity Press, Cambridge, 220pp., €60 euros (hardback) ISBN: 9781509549825
Many years ago, at the dawn of the digital revolution, I was invited to take part in one of the first public debates about the impact of mobile phones on everyday life. The panel was unanimous in deploring the use of these new means of communication. I particularly remember a distinguished philosopher, one of the European fathers of research into the ethical and social impact of new technologies, who was vehement in his condemnation. At the end of his talk, I wanted to join the eminent professor to thank him for his inspiring words. I could not, however, because when I got there, I found him already engaged in an intense and, from the tone of his voice, amusing mobile phone conversation on his colourful Motorola StarTac. I left the old sage to his witty conversation, peppered with accounts of academic shenanigans, and walked away, reflecting on the fickleness of the human mind. This episode always comes back to me as a cautionary tale when I am about to speak or write about the human impact of digital technologies. Indeed, one of the virtues of Anthony Elliott’s thought-provoking book, Algorithmic Intimacy, is its unconventional take on the world of digital communication, and its careful avoidance of moralism and hypocrisy.
Anthony Elliott is executive director of the Jean Monnet Centre of Excellence at the University of South Australia, where he is also research professor of sociology. His interest in digital culture and the material impact of the information revolution on human civilization dates back to the early 2000s. His background in the social and political sciences is evident when reading this essay, subtitled The Digital Revolution in Personal Relationship. However, being curious by nature, I took a look at Anthony Elliott’s curriculum vitae and discovered that, early in his career, one of his main interests was psychoanalysis. Psychoanalysis, or at least a view of society that is aware of the many complex layers of thought that make up the mind of the individual and that of communities, is the unspoken subtext of the whole book, which made reading it all the more fascinating and even intriguing to me. In saying this, I am declaring in advance my scientific conflict of interest: at a time when psychoanalysis seems to have been almost removed from the scene, I, as a clinical psychoanalyst, have a positive bias towards those scholars who have not. Although, as you will see, Anthony Elliott and I have our differences, Elliott speaks the same language as me. This is why the long-distance conversation I am about to begin with him, with you as participating readers, will at times resemble a fireside chat between friends (who do not know each other) rather than a traditional academic debate. I ask for your forgiveness in advance.
Intimacy and algorithms
‘“Algorithmic intimacy” as I define it’, Elliott writes, ‘has to do with advanced computational processes, known as machine intelligence, which produce new ways of ordering personal behaviour and modelling intimate relationships’ (p.7). Elliott devotes the entire second chapter of his book to an analysis of the meaning of the term ‘intimacy’, from Simmel to Bauman. It is mainly in these two authors that Elliott captures what seems to him to be the crux of the definition of intimacy, namely that it is a condition linked to the emergence, in industrial and modern society, of the idea of the private sphere. The subject – argues Simmel (1903) – is driven by the frenetic life of the metropolis to create a protective shell for himself, to shut himself in. The industrial economy produces an intimate, asocial type of human being. The loneliness of modern man is the urge that drives him to seek sharing, some form of communion with a few other people, so as to relieve the burden of the isolation in which he would otherwise live. In no different vein, Bauman (2003) describes the bonds of intimacy as almost the last resort of contemporary man, who is immersed, one might say jokingly, in liquid society like a goldfish in its aquarium. Intimacy creates both closeness and distance between the self and others, a differentiation that gives the person the feeling of existing as an ‘I’ in front of another who is not the person’s double, and provides an emotional security that allows the person to trust, to open up to others.
Elliott relies on this interpretation of intimacy because he finds it useful to show how algorithmic intimacy does not come out of nowhere but is the logical development of a condition that was already emerging during the industrial revolution. For Elliott, there is almost a common thread linking the loneliness of the metropolitan citizen described by Simmel to that of the Baumanian citizen in the liquid society, and finally to that of the digital citizen. If we try to express what Elliott does not say explicitly but hints at, we are faced with a process that follows its own internal logic, a progressive hybridization of human beings with the elements of an increasingly automated system of production:
In the algorithmic era of advanced automated societies, intimacy is coming to mean smartphones, social media, softbots, self-tracking technologies and the software-driven sharing of information. (p.34)
From this idea of algorithmic intimacy, Elliott creates a topography of intimacy in the digital age. He explores three distinct territories:
RELATIONSHIP TECH., concerned with the ways algorithmic conceptions of intimacy are influencing sexual relations, dating, marriage, family, eroticism and love. THERAPY TECH., concerned with the way algorithmic conceptions of personhood are influencing therapeutic mental health, specifically well-being, welfare and autonomy. FRIENDSHIP TECH., concerned with the way algorithmic conceptions of companionships are influencing mutuality, interpersonal bonds, communication and sociability. (p.17)
Elliott dedicates a chapter of his book to each of those domains. It would be almost impossible to discuss them in detail as they are so dense with information and analysis. Rather, I prefer to focus on three issues that I consider particularly important among those Elliott chooses to address. They are not the central themes of Elliott’s essay (in fact, they may seem marginal at times), but I do not think I am wrong in saying that they are an important part of the thinking behind this book. My suggestion to the reader is to approach Algorithmic Intimacy by also reading it through the lens of these three themes: (1) intimacy and magic; (2) intimacy and representation; (3) intimacy and shame.
Intimacy and magic
At the beginning of the section on relationship technology, Elliott mentions a crucial theme – the theme of magic. Indeed, the use of digital technology to bring people together and to create love and erotic relationships bears a striking resemblance to the ‘magic of love’ used for centuries by magicians and sorcerers to predict and often trigger love encounters through spells, filters and magical rituals. However, the relationship between algorithmic intimacy and magic goes beyond this and concerns three important aspects.
The first, directly related to the quantified universe described in Relationship Tech, concerns prediction. Only one other era has been as obsessed as ours with the idea of predicting future events – the Hellenistic civilization, steeped in Hermeticism, Orphic religion and Neoplatonism. Of course, every age has always been interested in the future in its own way. The ancient Romans were almost ridiculously superstitious and did nothing without first consulting soothsayers (they were also so cynical and disillusioned that they were the first to laugh at the compulsion). Our ancestors at the medieval and Renaissance courts considered astrology and horoscopes to be sciences worthy of great respect (even Newton considered himself an astrologer rather than an astronomer). The seventeenth and eighteenth centuries were centuries of trying to predict the outcome of fate (in roulette as in commerce) by means of a new science invented ad hoc, the calculus of probability. In short, each age had its own way of trying to know what was going to happen. But if you look at the examples I have just mentioned, you will see that in all of them the prediction of the future is based on theoretical, scientific constructions (albeit in forms that do not seem scientific to us today). The Roman haruspices followed very precise rules, codified point by point, and nothing was left to intuition or improvisation; the astrologers of the Middle Ages and the Renaissance were skilled mathematicians, their predictions had a geometric accuracy; the probability calculus of the seventeenth century was rigorously based on new and refined mathematical techniques. In almost all epochs before our own, predicting the future claimed to be a science governed by clear-cut principles. This is no longer the case. Predictive analytics does not try to predict by understanding the laws that govern the way the world is becoming. One of the main characteristics of data science is its ability to ‘predict without understanding’.
Conventional science seeks mechanistic explanations based on causality. Probabilistic science looks for statistical correlations and – once strong enough correlations have been found – tries to establish some kind of causality. Data science is still looking for correlations and patterns, but it is no longer interested in extracting any understanding from them, but rather in developing capabilities to prompt and drive action through the early detection of these patterns as they emerge. The magical aspect of the use of algorithms to influence sexual relations, dating, marriage, family, eroticism and love, therefore, lies not only in the opacity of the algorithms – as enlightened by Elliott – but also in predictions that are not based on any principle of causation (not even a ‘magical causation’), but on the mere repetition of patterns that are nonsensical in themselves. Not entirely as a joke, data science is closer to divination and prophecy than one might think.
A second aspect of magic is that of ‘presence’. To be present is to be ‘here’ and ‘now’. Presence is the notion that almost all forms of magic challenge. Magic tends by its very nature to be evocative; i.e., it aims to ‘call’ into the present what would naturally be absent (e.g., a departed soul, a spirit, a demon, a person who is in another place). This magic claim of making something or someone present is possible because ‘present’ is itself an oxymoron since everything necessarily flows in time and cannot ever really exist in a ‘here and now’. The radical impossibility of ‘presence’ was once challenged by magic. Today it is challenged by the internet. The presence/absence of the partner sharing intimacy plays a crucial role in the online erotic relationship as well as in the therapeutic relationship. Digital technology creates a strange hybrid of presence and absence. Elliott mentions it explicitly in the chapter on Therapy Tech, which I think is one of the most enjoyable and useful, both for the completeness of the picture it describes and for the subtle irony that emerges from many of the descriptions (incidentally, unless I’ve missed it, Elliott does not mention the concept of ‘transference’, which would need to be discussed when talking about online AI-based therapy).
The issue of presence also concerns representation. I will talk about representation in more detail later, when I address the question of the staging of intimacy. I mention it now because Elliott addresses online presence in a way that I found original and interesting, namely through the theory of the imaginary formulated by Cornelius Castoriadis (Elliott deserves special praise for remembering this brilliant philosopher and psychoanalyst and his theory of the imaginary, which is too often overlooked). What Elliott suggests, following Castoriadis’s lesson, is that the quantification of relationships is a way of exercising control over fantasies, especially sexual fantasies, and displacing the physical presence of others (see Elliott, 2002). Elliott’s hypothesis – with which I absolutely agree – is that ‘quantified relationships’ also aim to protect the subject from the fear of bodies. Like any psychological symptom, algorithmic intimacy serves two masters: on the one hand it is a tool to facilitate loving and erotic relationships, and on the other it is a means to ‘sterilize’ and disembody them.
Finally, a third aspect of algorithmic intimacy in relation to magic has to do with animism. Elliott mentions the process of attributing intentions, emotions, wills and desires to machines in various parts of the book, most notably in the section on friendship technology, where he addresses the vast topic of automated friendship, including robotics, virtual assistants and conversational agents. While Elliott is positive about many aspects of these technologies, he cannot help but conclude that the field of automated friendship poses two serious risks. A first risk is that automated friends may, paradoxically, increase feelings of loneliness and isolation. Although Elliott does not mention this explicitly, it is likely that he is thinking of the ‘uncanny valley’ effect associated with artificial intelligence (AI). An excessive similarity between machines and humans increases the distance we feel towards machines, their unresolvable otherness. There has been much debate in recent years about whether AI has leapt over the uncanny valley and is now completely indistinguishable from humans; Elliott is clearly not of this opinion (if only because a machine can be switched on or off by an operator) and this author is in complete agreement with him. A second risk is that the subject is enclosed in a solipsistic universe, in a narcissistic dream populated by pseudo-others who make no decisions and have no autonomous will. However, the problem of the relationship between man and machine, which has become a fictional replica of the relationship between human beings, does not end with these two risks. There is another important aspect, known as ‘techno-animism’.
Human beings have an innate tendency to regard inanimate objects as living and conscious (animism) and to ascribe human qualities to them (anthropomorphism), perhaps because they are animals that relate to the world through continuous processes of symbolization. The term ‘animism’ was coined by the English anthropologist Edward Burnett Tylor to describe the ‘earliest’ period of magical-religious thought. Tylor defined animism as the belief that animals, plants and inanimate objects all have souls, and attributed this phenomenon to dream experiences in which people often feel they exist independently of their bodies. For Tylor (1871), animism was a ‘Stone Age religion’ that still survived among some of the ‘ruder tribes’ the British encountered in such places as Africa and southern India. The child psychologist, Piaget, also found that the youngest children see virtually all phenomena simultaneously as alive, conscious and made by humans for human purposes (Piaget, 1962). In recent years, several authors (Guthrie, 1993; Bird-David, 1999; Ingold, 2000; Harvey, 2005) have shown a renewed interest in animism. Their view of animism differs significantly from the traditional definition. Rather than a ‘primitive’, ‘childish’ superstition that ascribes life to the lifeless, they see animism as an alternative response to universal semiotic anxieties about where and how to draw boundaries between people and things. It is precisely these boundaries that are threatened by new digital technologies. In their seminal book, The Media Equation, Reeves and Nass (1996) show that interactions between people and ICTs are identical to real social relationships. People automatically extrapolate personalities from small clues. In all human–machine interactions, personality can creep in everywhere: the language of error messages, user prompts, methods of navigating options and the fonts chosen. Marc Pesce (2000), one of the early pioneers of virtual reality, speaks of ‘techno-animism’ to describe a world permeated by computational objects. The algorithmic intimacy is making the boundaries between things and human beings even more blurred. We are now living in a world that is not only imagined to be saturated with invisible communication between us and the things that surround us, but which is in fact so. Digital objects can observe, learn, communicate and act in the real world.
We live in a world that bears an uncanny resemblance to the cave of the sorcerer’s apprentice in Walt Disney’s Fantasia, where Mickey Mouse, a young apprentice, tries out some of his master’s magic tricks, but has no idea how to control them. The worst fate for a fantasy is to become reality, or to be thought of as reality. The two realms – imagination and reality – must remain distinct for the sake of individual and collective sanity. I see the increasing confusion between them as one of the main risks of algorithmic intimacy.
Intimacy and representation
Intimacy is a complex notion, constructed as an inner space of secrecy from others in relation to evolving social norms. As a ‘private’ space, it deprives others of their possible control by removing something from their gaze. The space of intimacy is of variable geometry: what remains secret to one person is not secret to another, and boundaries evolve according to personal and relational conditions and times. Intimacy cannot be reduced to a mere register of what is private as opposed to what is public, nor can a sharp opposition be drawn between secrecy and transparency. The border between the private and the public is blurred and difficult to grasp because each individual, faced with social prescriptions, is constantly oscillating between hiding and showing, depending on the relational and institutional context.
Today, intimacy has become public, while still being defined as intimacy. It can even be said that intimacy is public precisely because it is intimate for otherwise it would lose public interest. This brings us to the paradoxical aspect of the evolution of our relationship with the expression of intimacy in the public sphere: on the one hand, transparency is the new social norm, while on the other, the protection of privacy has become a collective obsession. The term ‘privacy paradox’ refers to the curious fact that people who are accustomed to protecting their privacy in normal circumstances do not mind publishing almost anything about themselves online. Even people who know very well that the publication of their pictures online will cause them a cascade of problems (e.g., fugitives wanted by the police, mafia bosses, etc.) do not hesitate to publish their pictures on social media. Why does this happen? A first answer might be that people need to be recognized to feel that they ‘exist’. The internet, especially social media, is today the most powerful system for being recognized, and many people feel they exist only when and because they are online.
In this sense, it would be worth taking up Kenneth Duva Burke and his ‘dramatism’ (Burke, 1985) and Erving Goffman’s dramaturgical approach (see Brown, 2003). Guy Debord (1970), with his ‘society of the spectacle’, and Marshall McLuhan are also fundamental to understanding the ‘theatrical’ dimension of algorithmic intimacy. In McLuhan and Nevitt (1972), McLuhan speaks of the global electronic theatre (made up of the self-assemblage of all electronic media) in which all the world’s inhabitants are both actors and audiences. The repertoire of such a theatre consists of a perpetual happening, including the replay of every happening of the past. McLuhan’s view was ahead of its time. Indeed, the new global media system is turning the world into a global stage, where borders between communication, information and entertainment no longer exist, and everyone – willingly or unwillingly – plays to a global, digital audience (Aspling, 2011; De Kosnik, 2015). John Tinnell (2011) has developed the concept of the world browser, where:
the world serves as a stage for the performance, and because virtually any space on Earth can serve as a backdrop for live footage, people live with the awareness that they may become ‘actors’ in that footage at any moment – intentionally or not.
In the past (Mordini, 2011), I have come to define this as the concretization of the ‘Pulcinella secret’ on a global scale. Pulcinella is one of the oldest characters of the Commedia dell’Arte. He is the stereotype of the lazy, insolent and chauvinistic servant, sometimes stupid, sometimes clever, always penniless and absolutely incapable of keeping a secret. In a typical plot, the master tells Pulcinella a secret, swearing that the servant will never reveal it (usually a love secret). Ignoring the oath, Pulcinella immediately starts telling everyone the secret. Each time he does so, Pulcinella makes it clear that he would never reveal the secret except to that particular person, and in any case, he demands that the new recipient of the secret swear never to mention it to anyone. Needless to say, all those who receive Pulcinella’s secret take it in turns to reveal it, repeating Pulcinella’s arguments (i.e., first they declare that they will reveal the secret to that person and only that person, by way of exception; then they impose an oath not to reveal the secret to anyone else). Thus, in a short time, all the characters on stage know the secret, but none of them knows that the others know it too. On the contrary, they all pretend to ignore it. This situation is used to create a series of misunderstandings and comic situations. The moral of the story is that secrets are almost never secret, but rather social conventions that it is convenient for everyone to pretend to believe. Something similar seems to be happening with intimacy today.
Intimacy and modesty
I said that intimacy implies the delimitation of an inner space, a discontinuity against a background of continuity between me and the other, which is the opposite of confusion, transparency and appropriation. It presupposes a delimited and preserved space, an exclusion, a secret part reserved for oneself and for those we call ‘intimate’; that is, those with whom sharing is possible and authorized. It requires the existence of a boundary, a dividing line between what is shown and hidden, veiled and unveiled, visible, narratable, audible or not. At the same time, intimacy is not a purely intrapsychic process of the solitary self, but necessarily unfolds in intersubjectivity; in a sense, it is always public, since it concerns that which is hidden from the sight and knowledge of others. When Elliott repeatedly refers to the fact that new technologies create new social practices, new collective paradigms, new values and habits, he is referring precisely to this intersubjective dimension of intimacy. Intimacy exists only when it is shared and recognized by another who respects it.
The idea of respecting intimacy is integral to the concept of intimacy. To question intimacy is to question modesty and shame. Through its concealment, modesty reveals the preciousness of what is kept secret and thus the value of the revelation in which only some can be participants. Modesty is defended mainly by two virtues or behaviours: tact and discretion. A lack of tact and discretion can lead to a feeling of shame in the ‘victim’. Shame comes from having revealed something one wished to keep private. At the extreme end of shame is the experience of humiliation, which always involves some form of degradation. The term ‘humiliation’ is etymologically derived from the Latin word ‘humus’, meaning earth. It refers to the situation in which, during a fight, one of the two combatants is forced to lie face down in the dust or mud. All victims of severe humiliation agree that one of the most distressing elements of this experience is that it inevitably involves some element of self-abasement. The humiliated person could have resisted being humiliated, but there was often a very high price to be paid for rebellion, in some cases life itself. One accepts submission to humiliation because one does not have the courage to rebel and pay the price of one’s rebellion. Robert Antelme accurately described how resistance to the dehumanization of the Nazi extermination camps was expressed through the preservation of a place of modesty, the latrines (see de Montleau, 2010). Other camp survivors, such as Primo Levi, showed how the obligation to wash in front of everyone in the lead wagons was the first act in a process of increasing degradation.
The process of self-degradation that humiliation entails can unfold in a number of different ways, not necessarily in the explicit ways that have just been mentioned. In the world of algorithmic intimacy, it often follows a specific dynamic. The digital world today seems hungry for intimacy, of which it captures all the most emotionally engaging aspects, albeit at a superficial level. Social media, reality shows and so on create an atmosphere of collective intimacy that seeks to show everyone’s dirtiest, most vulgar aspects rather than their best. This ludicrous parody of intimacy allows everyone to identify with a set of stereotypes that often corresponds to as many consumer niches. By overcoming modesty and shame, by removing all forms of discretion and tact, collective digital intimacy becomes a tool for creating individuals who are apparently different, but are in fact all the same, like different models of the same smartphone. It is an intimacy that, far from revealing what it means to be a person, accentuates anonymity. Algorithmic intimacy, understood as a crude intimacy that ignores shame and modesty, can become an instrument of cultural alienation, a means of promoting that process Pasolini called the ‘homogenisation’ of our societies (Garguilo, 2022).
What we are
The book ends with a brief description of three cognitive strategies, in the sense in which Habermas uses this term (see Filipiak, 2017), with which to approach the world of algorithmic intimacy. These three strategies – conventional algorithmic intimacy, cohesive algorithmic intimacy, individualized algorithmic intimacy – convincingly describe various responses to the different situations outlined in the book. The first strategy is based on the possibility of using algorithms to generate structured hierarchies, processes and rules of operation; the second strategy aims to create communities and relationships between people; the third aims to produce a digital imaginary that privileges the dimension of imagination, creativity and communication. I could discuss them in detail, but their linearity and clarity do not require much comment. As with the previous chapters, I would like to propose my own key to this final chapter. For this, I will start from the last paragraph of the book, which is dedicated to a quote by the poet Paul Valéry: ‘We never think that what we think conceals what we are’. Like those mystery novels where there is a revelation at the end that forces readers to change their perspective and reinterpret the meaning of what they have read, this quotation forces us to rethink Elliott’s essay.
Up to now, we have thought of intimacy as being made up of meanings that can be shared with another person, who becomes the repository of them in the course of confidences. Now, at the very end of the book, Elliott reminds us, through the medium of a poet, that there is a side of intimacy that is secret to ourselves, revealed precisely because we cannot grasp it. Intimacy contains a bottomless pit that eludes us and about which we cannot speak directly: the mystery of our otherness. This lack of knowledge of the most intimate part of ourselves is sometimes revealed publicly without our being aware of it, through our facial expressions, our posture, the redness of our face, the tightness of our skin or when we say more than we mean to, by ‘accident’. It is the effect of surprise (when we blush, for example) that reveals this strangeness to us. This register of unspoken intimacy is revealed more in these ‘accidents’ than in situations of intimacy, although they can also occur in a confidential relationship. In everyday life, it is the people close to the person, or more precisely their intimates, who know more about the person’s intimate background than the person does. A psychoanalyst might say that this kind of intimacy, unknown to the subject but vaguely perceived by him, coincides with the unconscious; this would be a partly true statement, but it would still not capture the deeper meaning of this experience. This side of intimacy represents what we know about ourselves without being able to express it, the insoluble riddle within each of us. It is an enigma that cannot be fully unravelled, not even by AI. There is a story, a mysterious and fascinating story, that can be taken as a symbol of this enigma.
In ancient Athens there was a religious ceremony called Arretophoria, whose name literally means ‘transport of unnamed things’. It was an all-female festival in which men were not allowed to participate. Two girls, aged between 7 and 11, were chosen to live for a time on the Acropolis, in the temple of the goddess Athena. The young girls wore white robes decorated with gold and ate a special food called anastatos. On the night of the last full moon of the year, which fell in summer according to the Athenian calendar, these girls were each given a wicker basket in which the goddess’s priestess placed some veiled objects. Neither the priestess nor the girls knew what they were. The girls carried the baskets on their heads a short distance to the gardens dedicated to Aphrodite, the goddess of love. Here they would enter a cave, leave the two baskets and take two unopened baskets of unknown contents. They took these two new baskets to the temple on the Acropolis and the ceremony was over. It has never been known what the baskets contained, either on the way there or on the way back. Some scholars believe it was grain paste or stones depicting snakes, phalluses and human forms, but this has never been proven. Others have suggested that the baskets were simply empty (Cruccas, 2007).
The Arretophoria, a ceremony structured around an ineffable mystery, was one of the most important religious rites in Athens. Its meaning – which remains largely unknown to us – seems to refer to the biological and social processes entrusted to intimacy that underpin the order of the polis, such as marriage, procreation and the education of the young (Robertson, 1983). If this hypothesis is correct, then the Arretophoria, by staging the mystery of female intimacy, would represent the realization of human order in opposition to the chaos of nature. Finally, intimacy would emerge as one of the main organizing forces of human civilization.
I think Anthony Elliott would agree.