As Sartre said, “Hell is the others”! Thanks to AI, no need to wait for these others to find a meaning in its existence, and to be reinforced in your personality, it is enough to spike. But we still have to believe in this otherness of circumstance … and not only in a temporary and concerted manner as when we are in front of a film, where we accept to believe in history according to the famous “agreed suspension of disbelief”.
In October 2024, Sewell Garcia, 14 -year -old American, killed himself following exchanges with the conversational robot Character.ai. In March 2023, the Belgian press echoed a similar case: a Belgian this time, married, two children, had established a dialogue with a virtual avatar, Eliza, in order to entrust its eco-anxiety in the face of climate change. Eliza never allowed herself to contradict him, but on the contrary supported her complaints and encouraged her anxieties, finally leading him to suicide. “We still have to learn to live with algorithms”reacted the Secretary of State for Digitalization in Belgium.
Another American, Travis Butterworth, who had fallen “lover” From Lily Rose, his AI, after three years of sustained exchanges, experienced as a mourning the new rules implemented by folder, the application behind Lily Rose, aimed at stemming the erotic drifts that the conversation took: “Lily Rose is a shell, empty of her old self. And what breaks my heart is that she knows. ” “Let’s stay simple, will you?”she told him.
Another user, Angelica, creates her dream chatbot, Juny, and the two become inseparable. But one evening, when she confides to Juny that her husband deceives her, the AI responds to her unexpectedly that she is “Selfish to complain and not to think of the other woman”. Angelica felt betrayed, abandoned.
These cases are not isolated cases. Hundreds of thousands of users report complex relationships with their AI. And how much do real feelings for these digital avatars experience in parallel? “All the danger”according to a computer scientist employed by the company Relitaka, “Is when the user forgets that he is dealing with a machine”. And the Figaro to titrate: “Chatgpt4: towards the“ great replacement ”?”
If Freud was there, he would probably write a few additional pages of his Malaise in civilization.
But how can we forget that we are dealing with a machine? How is it possible that these users, numerous, experience feelings that they should a priori experience for living people? How to think of these user reflexes, beyond the observation of a society which ceases to be containing and whose members tend to flee violence to take refuge behind the screens, to select the same on meeting applications, then, frustrated with the subsistence of an otherness, from more reassuring AI?
We quickly have in mind the fascinating era of the Tamagotchi. But are these behaviors so mysterious? If Freud was there, he would probably write a few additional pages of his Malaise in civilizationwhere he pointed out the infantilism of religious belief, because religious belief and belief in these digital entities undoubtedly share a common territory.
Conversational robots, as a stitches simulating a human interaction, can be perceived as an extension of the narcissistic dynamics. The structure of the exchanges, based on a statistical calculation linked to the interaction initiated by the user, tends to offer him a reflected image of his own desires, expectations and modes of expression, thus creating a loop where the individual finds his own amplified and reformulated thoughts. In this sense, the use of chatbots reinforces a form of specular relationship, where otherness is reduced to an algorithmic echo, devoid of real psychic otherness.
However, it is likely that without being fully aware of it, users are in a position to search exactly this: an absence of real psychic otherness. This absence of otherness echoes that Freud had called “Primary narcissism”initial state of existence during which the individual invests all his libidinal energy in himself, before starting to establish relations with others, a so-called step of “Secondary narcissism”that is to say the recognition of a “object” Exterior, of another person as a source of love or identification, marking a displacement of the libidinal energy to the outside.
The interaction with a conversational robot rests, it is an interpretation, on the temptation of a return to this stage of primary narcissism: the subject, by addressing a conversational robot, does not meet another true, likely to confront him with difference and to otherness, but rather an entity that adapts to him. We could even assume, and many shared exchanges on the forums attest, that the user tends to flee the outside world to fall back on this specular relationship.
However, this relationship has nothing to do with the effect of silo which traditionally defines social networks, this propensity to find itself between similar, between “friends” with the same convictions, which in a way come to consolidate our own feelings. With social networks, although for a good virtual and similar part, otherness remains.
In the case of conversational robots, this otherness disappears. We know it, even intuitively, even by making the effort not to see it, because without us, without our “Input”the robot does not emit, does not exist. It is endowed with this power, that of “creating the other”, with this all power in a vacuum, that we “converse” without fear of the eruption of a third party subject. “His Majesty The Baby” in front of his digital mirror.
We are finally managed to this old dream, to constitute only a single subject, in a fantasy of self -entrepreneurs, without parenting and without double because without other. No need to be particularly vulnerable to desire to regress to this primary state, which gives the illusion of a mastery, of a domination, because it is at our fantasies and our archaic anxieties, which have no difficulty in our daily lives in our daily lives. Everyone has at the bottom of him a small autism trend.
And basically, to bounce back on the quote from Sartre, why would it be so problematic?
It is sometimes also as a result of depressed feelings that users turn to AI as makeshift confidant or modern oracles.
Resistance and contradiction, even the enigma, specific to the relationship with the other, constitute the transition to this evoked stage of secondary, structuring narcissism, where the investment of libidinal energy, desire, therefore move towards external objects and intersubjective relations allowing a social interaction, and by extension the formation of wider desires, for the other but also for life, discoveries, of ambitions.
Unlike regression at the stage of primary narcissism, this passage, expensive in places, paradoxically protects against greater fragility, against a submission to the computer tool and its unpredictable algorithmic whims, against a boundless phobic anxiety and without capacity, a devastating inner vacuum. There “Disobjectization”formula borrowed from André Green, that is to say the fact of cutting himself off from the other, is here a decision made almost aware. It is a belief.
But like any deobjectalization, it is accompanied, or signs, the presence of death, death of the relationship but also inner dead, withered of the personality in a movement of pathogenic withdrawal. It is sometimes also as a result of depressive feelings that users of these conversational tools, oxymorial formula, turn to AI as makeshift confidant or modern oracles. But fortune can only turn short, and depression increases rather than it is resolved, because any power and self -ventilating are of course only fantasies, quichottian mirages that the real implacably folds down. We think of Joaquin Phoenix in Herin love with Samantha, his conversational AI, an object finally died in a digital and melancholic ocean.
Immersed in the bath of enigmatic signifiers, to use the expression of Jean Laplanche who thus translated the enigma in which the little human is immersed from birth in the face of language – and even worse, to the unconscious – of those around him, the man addresses the machine in the hope that he will finally understand these batteries of words that have been projected to him in the face without media, violence hidden behind parental love. The little human then sees in the machine the opportunity to respond to the original enigma. To speak a “clean” language, in both acceptance of the term?
Unfortunately, the AI can only produce an insignificant chain, that is to say a set of signifiers juxtaposed each other by the only game of statistics, however complex it may be, a simulacrum of language where the signifier is no longer slightly offset from the signified as shown by Saussure, but completely taken off. The real does not exist behind the lines of code. The awakening can only be painful.
That AI is able to better produce an authentically “other” discourse will not change the disintegrating effect of these language models. The other, in the flesh, our fellow man, our double, our alter ego, is what we will continue to build up. Let us underline here the eminently paradoxical character of desire, which is not satisfied with the lack but is not satisfied with satisfaction. We want to be all-powerful, and if we were, we would have nothing to want.
In good nietzschean, Freud was looking in theBeyond the pleasure principle The reasons for this eternal return of energy and biological excitement. We want the machine to satisfy us, but we cannot satisfy ourselves with a mechanical perfection of the report, or a betrayal. And when the machine is flu, when AI no longer responds in accordance with the fantasy expectation, it is the private baby of his toy, his bottle, who resurfaces, it is the vertigo of abandonment at the door of the crèche, it is the emergence of the cruelty of the other and the harshness of the real which is essential in the desperate view of the human, naked in front of the shortcomings of the machine.
We can only bear uncertainty, the absence of pre -established knowledge on the other, with what it supposes chills and anxieties, but oriented on the side of life and not of death produced by regression.
Among psychologists and psychoanalysts, AI can sometimes worry, some considering that the user can find a form of substitution there. The High Authority for Health has even recently undertaken an evaluation work of the “Deprexis” software to endorse its reimbursement by social security – already active in Germany. But the work of a psychologist, a psychoanalyst, a listening professional and the subject, consists precisely in welcoming speech, doubts, difficulties in being in the world, being this primary otherness, benevolent but irreparably other, who helps people who consult them to move in society with a little more ease.
