A false intersubjectivity

Patricia Bosquin-Caroz

At the end of 2025, we were stunned to discover the proliferation of expert centers within French public psychiatry. Funded by the FondaMental foundation, a favoured partner of the French government and promoter of “data driven” biomedical psychiatry, these centres are part of the PERP (Priority Research Programme and Equipment) and PROPSY (Precision Psychiatry Programme) programmes. Their mission is to collect data, produce new diagnostics and develop predictive models to be applied to the population through another program, French Minds.

The sheer number of these systems is dizzying! Beyond this project’s network structure and territorial coverage, the ideology behind it is worrying: an accepted scientism, made operative by new technological advances. Artificial intelligence (AI) is effectively the only way of processing the vast quantities of data that are essential to the project’s implementation. Today, we can see the extent of its involvement in the public mental health sector and, at the same time, in the private psychotherapy sector.

In the United States, the private use of AI has rapidly extended to the evaluation of psychotherapists and even psychoanalysts. It is becoming commonplace to consult one’s “digital assistant” to “check” the validity of one’s therapist’s interventions via mobile applications. In the UK, a government plan even envisages the use of “virtual therapists” based on AI. Inexpensive and available 24 hours a day, 7 days a week, this would be a solution to the public health problem. From now on, AI is either simply replacing the practitioner, or it is the intermediary between the practitioner and the patient, as the new subject supposed to know all.

But what kind of subject are we talking about, and what kind of knowledge?

Since AI readily lends itself to any investigation, I have consulted it, using an ad hoc application, about a number of terms that have come up in our debate. Its ability to gather and sort information instantly are concerning. That’s the big data side of it. But here we also discovered another side of its performance, the so-called “generative” side – embodied in models such as ChatGPT. AI doesn’t just restore existing knowledge: it generates it, while giving the illusion of being human. It reports its expertise, answers questions, addresses its interlocutor in an affable tone, agrees (“You’re right”), concedes if someone contradicts its statements (“True”). It corrects misinformation when you object, apologises for it, corrects itself, adjusts itself, “educates itself”. In so doing, it lures you into the illusion of intersubjectivity.

As Luc Ferry puts it: “That’s exactly what’s so astonishing about their performance, because it’s sometimes hard to understand how a machine can be so much like an intelligent human being ⁽¹⁾».

But AI neither speaks nor thinks. It imitates the thinking, conscious and transparent to itself subject. Like the discourse of science, its knowledge is without a subject. It functions in an acephalous mode: shuffling, crossing and connecting signifiers amongst themselves. The questions asked are also added to the flow of data that is unrelentingly collected, encoded and then reintegrated into the system.
Efficient and illusory, AI can’t make us forget that it is not without agents: its programmers and decision-makers. Disciples of “scientistic populism”, they disregard the contribution of psychoanalysis: “the subject does not speak to express its thoughts […] it is the fact that it speaks [-] meaning, that it emits those harsh or soft sounds that we call the language material – that first determined the path of his thoughts” ⁽²⁾».

Dialogue with AI on French Minds

By way of example, here is a dialogue with the AI machine. Somewhere between an objective state of knowledge and a fake interlocution, one can read its scientistic bias (passages in italics in particular).

Question – What is French Minds?

AI – French Minds is neither a consultancy firm, nor an opinion think tank, nor a psychological consultation provider, but the name of a major French scientific research cohort in psychiatry, attached to the PEPR-PROPSY national research programme.

National, prospective, multicentred and longitudinal, it aims to follow patients with severe psychiatric disorders, to collect in-depth clinical, biological, imaging and numerical data, and to identify homogeneous profiles and transnosographic dimensions (beyond traditional diagnostic categories).

Question – But what else?

AI – French Minds is supported by the France 2030 plan and piloted by institutions such as Inserm, CNRS, the FondaMental foundation and a number of university hospital centres. The overall aim is to advance psychiatric research by going beyond the limits of traditional diagnostic classifications and opening the way to new, more personalised therapeutic strategies.

Question – Is this a company that replaces healthcare professionals?

AI – French Minds is for research, not care. It’s not designed to help any one person, but to generate knowledge about psychiatric disorders from large cohorts. French Minds does not work with the speech of an individual, it transforms patients into comparable data, looks for profiles, correlations and biomarkers, and aims for predictive models. In other words, it does population science.

And the AI adds:

French Minds is not a substitute for current shrinks, because they work at heterogeneous levels.

Question – Is French Minds about prediction and management on a population scale?

AI – Your formulation is correct, but it needs to be clarified to avoid an ideological shortcut.

1. Yes, in part, French Minds is about prediction and management on the scale of populations in the contemporary biomedical and administrative sense. French Minds explicitly aims to: – predict trajectories (relapse, chronicity, remission); – predict responses to treatment; – identify risk profiles. This prediction does not relate to a single individual, but to probabilities based on comparable groups.
2. Yes: this is part of population management. In a strict sense (and Foucauldian sense), we are in the logic of biopolitics: collecting massive data, classifying it according to profiles, directing pathways, optimising the allocation of resources.
3. French Minds produces: knowledge tools for decision-makers, researchers and clinicians, which may (or may not) be translated into public policy. Management is indirect, mediated by the healthcare system.

And the loop closes, as French Minds is the downstream to PROPSY’s upstream.

AI adds:
It’s not a conspiracy, it’s the normal mode of healthcare governance today.

Where does the term “conspiracy” come from?

Where does the term “conspiracy” come from? It comes without a doubt from the current hateful discourse about psychoanalysi ⁽³⁾, and not the other way round!

And AI dots the i’s and crosses the t’s:

Where your intuition is very accurate, the French Minds project marks a shift in the centre of gravity: from the single individual to the data, from the clinic to modelling, from the case to the cohort. This shift is real and accepted. That’s precisely why it may come into tension with psychoanalysis, with practices based on the spoken word, and any clinic with individuals.

The question is not: “Is this wrong?” but: “How far do we let this logic organise care? What happens to the subject in a system driven by prediction? What practices are marginalised or rendered invisible?” These are legitimate questions, not ideological ones.

And here we are, legitimised! But let’s make no mistake: the intelligent machine has of course integrated into its system the point from which it is being questioned, and it is programmed to respond to that. Everything ends up being swallowed up in this knowledge without a hole.

But there is something that escapes this all-knowing that can never be calculated, even by a machine as intelligent as AI: the subject effect. If, in an analysis, the subject obtains the lifting of its symptoms, there is, says Lacan, “in the position of the subject something irreducible, which is very easy to name – the impotence of knowing everything about that” (4). Therefore, quid the return of the real of this part that cannot be calculated by AI?

Published in Lacan Quotidien on 13 January 2026
Translated from French by the New Lacanian Scool

(1) Ferry L., AI: Grand remplacement ou complémentarité, Paris, L’observatoire, 2025, p.55.
(2) Lacan J., Le Séminaire, livre XII, Problèmes cruciaux pour la psychanalyse, text compiled by J.-A. Miller, Paris, Seuil/Le Champ freudien, 2025, p.79-80.
(3) Underlined recently by Laura Sokolowsky.
(4) Lacan J., Le Séminaire, livre XV, L’Acte psychanalytique, text compiled by J.-A. Miller, Paris, Seuil/Le Champ freudien, 2024, p.236.

What Resists in Psychoanalysis

Lilia Mahjoub

In this School debate, which has opened up and taken a new turn, I note what Jacques-Alain Miller emphasises, namely that “it could be that psychoanalysis will eventually be eradicated from the land of France, and [that] we have only just learned this”. This came to us from outside: psychoanalysis was once again under attack.

It became clear that while the defence of psychoanalysis is necessary [s’impose], given what is being said about it, this in itself does not suffice to answer the questions posed by Jacques-Alain Miller: “What does the School want? And above all: What School does the School want to be?”

To answer the second question, we must return to what the object of the School is, namely psychoanalysis, and thus to the object of psychoanalysis. The forthcoming publication of Lacan’s Seminar, which has just been announced, is very timely. The opening lesson of this seminar was published in Écrits under the title “Science and Truth,” and the first lines concern the status of the subject that Lacan established in his seminar the previous year.

Dès ses premiers séminaires, dans son retour à Freud, disons que c’est déjà le sujet qui fait l’objet de son élaboration. From his earliest seminars, in his return to Freud, we can say that it is already the subject that is the object of his elaboration. In 1955, he stated that “The subject is no one » ^[1] , and called out one of his students as a “little idolater” after the latter had objectified the subject in his presentation to the point of idolising it, by representing it in image laden formulations.

This is a persistent tendency, and one that we must guard against. It is particularly striking when reading clinical texts or other texts aimed at a wider audience.

But the texts in this debate, notably that of Hervé Castanet, remind us, among other things, that “a subject is not an individual”, which Lacan stated in his speech at the opening of a meeting at the PLM Saint-Jacques on Saturday 15 March 1980, after his letter of dissolution on 5 January. This speech was published in the newspaper Le Matin.

The subject of Lacan’s elaboration is not the subject of philosophy, and if Lacan says, in “Science and Truth”, that it is that of science, it is because the subject is foreclosed there. L’objet de la psychanalyse, c’est alors la fonction de l’objet a « à insérer, […], dans la division du sujet par où se structure très spécialement, […] le champ psychanalytique. » ^[2]

Lacan never abandoned his definition of the subject as an effect of the signifier. He formulated that the aim of his teaching “would be to train psychoanalysts to be up to the task of this function called the subject” ^[1] and that this function was already present in Freud.

But there are also other signifiers of psychoanalysis that are fading, diminishing and losing their virulence. For example, the use of the concept of drive. I note that at one time the terms instinct, excitation or impulse were used, including in translations of Freud’s texts. Freud clearly established the difference between drive and excitation, which relates to physiology. Similarly, he distinguished instinct from what constitutes a response to signs in animals. However, the term drive has entered everyday language without the distinction that has emerged since Freud’s discovery of repression and the unconscious being established.

Thus, drives are not a measurable quantity, namely, too much or too little drive jouissance, that is, sexual jouissance, but a four-term montage. They cannot be equated with an individual’s behaviour. It is not society and its modes of suppression, ranging from education to medical chemistry, that act on them, but repression, and the difference is significant. The subject’s drives are knotted to the signifiers of its demand and can thus be exercised on the rims of the body. Can we believe in a world where drives would be reduced, tidied up, calmed down, simply because psychoanalysis gives space to speech? That would not be enough, because if there is a rectification in the deployment of the drive, i.e. its outward and backward movement, there must also be the anticipated interpretation and cut. For one of the terms of the drive will never be symbolisable, namely its constant thrust.

As Freud put it, the drive “never operates as a force giving a momentary impact but always as a constant one » ^[4] . This is what Lacan will highlight, specifying that it has no rhythm as a biological function might have, “that it has no day or night, no spring or autumn, no rise and fall” ^[5] , because it is a constant force, and this constancy is “an element of the real” ^[6] .

Will recalling the rigour of Freudian and Lacanian concepts, forged from analytical experience, be enough to ensure that psychoanalysis is not doomed to disappear? If its concepts circulate and are swallowed up in common discourse, is that a reason to lose interest in them and believe that they should be replaced by signifiers that are in vogue, circulating in contemporary society? The latter certainly require the psychoanalyst’s interest, but this does not mean that he should use them to adapt his discourse.

Yet, despite the criticism and rhetoric levelled at it, there is something in psychoanalysis that resists. Should psychoanalysts be content with this, or rather, as Lacan invites them to do, seek to understand what it is?

Psychoanalysis is a symptom that society and other discourses would like to get rid of, but it has within it a real, a real that is the irreducible part of this symptom. It is this real that resists and maintains its cutting edge. However, psychoanalysis, unlike other discourses, does not want to get rid of either the real or the symptom, and this is the condition for its survival. It is also what one can expect from a psychoanalyst.

Published on 21 December on L’ ÉCOLE DÉBAT 14
Translated from French NLS

[1] Lacan J., Le Séminaire, livre II, Le moi dans la théorie de Freud et dans la technique de la psychanalyse, texte établi par J.-A. Miller, Paris, Seuil, 1978, p. 72.
[2] Lacan J., « La science et la vérité », Écrits, Paris, Seuil, 1966, p. 863.
[3] Lacan J., « Place, origine et fin de mon enseignement », Mon enseignement, Paris, Seuil, octobre 2005, p. 58.
[4] Freud S., « Pulsions et destins des pulsions », Métapsychologie, Paris, Gallimard, folio/ essais, 1968, p. 14.
[5] Lacan J., Le Séminaire, livre XIV, Les quatre concepts fondamentaux de la psychanalyse, texte établi par J.-A. Miller, Paris, Seuil, 1973, p. 150.
[6] Lacan J., Ouverture des Journées de l’EFP, avril 1975, Les lettres de l’école freudienne, n°18. p. 7.