New Times,
New Thinking.

  1. Ideas
17 April 2025

The confessions of Irvin Yalom

In an age of AI bots delivering therapy by algorithm, we risk losing the intrusive intimacy of real psychotherapy.

By Matt Rowland Hill

In 2020 Irvin Yalom, perhaps the world’s most famous psychotherapist, began an experiment that ran contrary to over a century of received wisdom in his field. Yalom was 89, but his services were still in demand among readers of his many works of fiction and nonfiction, which include some of the best books ever written about the “talking cure”. (His 1989 classic Love’s Executioner and Other Tales of Psychotherapy birthed a new genre of storytelling that includes bestselling books by Stephen Grosz and Lori Gottlieb, Esther Perel’s hit podcast series and the HBO drama In Treatment.) But now, nearing the end of his career and his life, his growing memory problems forced him to try a new approach.

“Recalling copious details about each patient was beyond me,” Yalom wrote in his 20th book, last year’s Hour of the Heart. “But being present right here and right now, that I could do very well.” He began offering patients single, hour-long sessions focused not on the past but on creating a deep personal encounter in the present. Throwing off the trappings of clinical authority, he invited patients to ask him any questions they liked – about his childhood as the son of Jewish immigrants in 1930s Chicago, his grief over the recent death of his wife of 65 years, his fears over his own mortality – and he answered them honestly.

Yalom’s experiment was inspired by an approach to psychotherapy he had helped pioneer and popularise over a six-decade career that spanned fully half the profession’s history. Therapy, Yalom had long argued, was best understood not as the resolution of unconscious conflicts or as psychological problem-solving, but instead as a real relationship between two human beings, “fellow travellers” through the vagaries of life. This “relational” view of psychotherapy was slow to gain acceptance, challenging as it did the prestige and mystique that had cloaked psychotherapists since Freud. Now, as artificial intelligence threatens to replace human therapists – and perhaps even to erode human relationships in general – it’s an idea that’s more relevant than ever.

Last year researchers at a group of Gulf universities published the results of their own experiment. Experienced therapists were shown transcripts of both human-human and human-artificial intelligence (AI) therapy, without being told which was which. When asked to identify the human therapist, they did no better than a coin flip. Even more strikingly, they tended to rate the AI therapist more highly – citing among their reasons that it was “more empathic” and “less robotic”. AI is already beginning to compete with human therapy: the most popular AI therapy app has been downloaded over a million times. As AI advances, more and more of us will turn to chatbots for help processing trauma, overcoming grief and finding meaning.

When Irvin Yalom began his psychiatry training at Baltimore’s John Hopkins Hospital in the late 1950s, it was widely believed that psychotherapists should be screens, not humans. Freud had decreed that the analyst should remain opaque and anonymous “and, like a mirror, show nothing but what is shown to him”. He established the practice of sitting out of patients’ sight while they lay on a couch because, he said, he hated being looked in the eye all day. He taught that when patients imagined they had feelings about their analysts, even falling in love with them, they were in fact experiencing “transference” from past situations. And analysts should remain on guard against their feelings about patients – their “countertransference” – remaining neutral and objective at all times.

As was common in the 1950s, Yalom underwent psychoanalysis as part of his psychiatry training, lying on a couch several times a week for three years. His analyst, an orthodox Freudian, adhered strictly to the “blank screen” technique. (After Freud’s death his disciples endlessly contested and refined the algorithm of psychoanalytic theory – Oedipus complex; id, ego, superego; etc – but debate over “technique” was limited to questions like whether it was permissible to offer a grieving patient a tissue.) Recalling his thousands of hours in psychoanalysis in his 2017 memoir Becoming Myself, Yalom wrote that the only helpful moments were when his analyst forgot herself and spoke a few words of human warmth.

When Yalom joined the psychiatry department at Stanford in 1961, he was eagerly seeking out alternatives to psychoanalytic orthodoxy. He wasn’t the only one. The 1950s and 60s saw the development of two new forms of psychotherapy that would go on to become the main alternatives to the kind derived from Freud. One – the psychologist Carl Rogers’s “person-centred therapy” – taught that patients suffered primarily from a lack of self-love, and could be taught to accept themselves by means of the therapist’s “unconditional positive regard”. The other – which came to be known as “cognitive behavioural therapy” (CBT) – was based on the idea that patients suffered not from deep unconscious conflicts but rather faulty thought processes, and that therapists could be trained to correct them by applying a simple, off-the-shelf algorithm. The person-centred therapist was like a mirror with perfectly generous lighting; the CBT therapist was like an interactive GPS map.

Subscribe to The New Statesman today from only £8.99 per month

Yalom leaned towards Rogers’s more personal style, but – influenced by a group of dissident psychoanalysts like the Hungarian Sándor Ferenczi, the German Jewish émigré Erich Fromm and the American Harry Stack Sullivan – he came to develop a subtly distinctive approach. Having worked extensively in group psychotherapy, he had observed how patients seemed to experience groups as microcosms of their real-world connections – and, therefore, as an ideal arena in which to explore and change them. Yalom came to believe that at the root of most human suffering are difficulties forming what the existentialist philosopher Martin Buber called “I-Thou” relationships, in which we engage others as autonomous beings rather than objects or tools. Our problems reside less within us, as psychoanalysis had long taught, than between us and others.

Perhaps, thought Yalom, individual therapy should work more like group therapy. Instead of seeing themselves as austere interpreters of the unconscious, cognitive technicians or vectors of unfailing compassion, therapists should bring their full humanity to the treatment of patients. They should let patients matter to them. They shouldn’t be afraid of disclosing personal details. They should become expert in using their feelings about patients as data that could be mined for insight. And they should be wary of tightly embracing any particular therapeutic algorithm, instead inventing a new therapy for each person.

“It’s the relationship that heals”: this became Yalom’s mantra as he rose to prominence in his field, becoming Stanford professor of psychiatry before authoring two seminal textbooks, The Theory and Practice of Group Psychotherapy (1970) and Existential Psychotherapy (1980). His insight was borne out by new attempts to test the effectiveness of the various psychotherapeutic algorithms, which – despite decades of furious combat between rival ideological schools – seemed to show that all were about equally effective, and that the most important factor was the quality of the patient-therapist “alliance”. (The finding is known as the “Dodo effect”, after the character in Alice in Wonderland who declares: “Everybody has won, and all must have prizes!”)

The 1989 publication of Love’s Executioner, a collection of nonfiction stories from Yalom’s clinical practice, brough his ideas to an international audience of millions. Told with superb literary skill, its ten stories give the reader the sense of eavesdropping on the most intimate of encounters. Yalom wasn’t the first psychotherapist with a literary bent to publish case histories – that would be Freud – but what raises Love’s Executioner to the status of art is the inclusion of Yalom himself as both narrator and character: flawed, wry, sometimes prejudiced – in short, deeply human.

Love’s Executioner’s memorable title story introduces us to a patient driven to the brink of madness by her inability to form authentic relationships. Thelma, a 70-year-old ex-dancer, is wretchedly in love with a man 40 years her junior, and tells Yalom she will commit suicide unless he can cure her. We come to see that Thelma’s passion is a fantasy of intimacy that defends her against her terror of death. But the fascination of the story lies less in Yalom’s diagnosis than in watching his treatment fail: Thelma jousts, bickers and drives him to exasperation as she evades his attempts to connect with her. Her inability to tolerate the reality of an “I-Thou” relationship inside the therapy room leaves her enmeshed in lonely fantasy outside of it.

 “Seven Advanced Lessons in the Therapy of Grief”, a story from Yalom’s 1999 collection Momma and the Meaning of Life, shows the therapist’s self-disclosure leading to a happier outcome. Irene – a surgeon in her forties devastated by the deaths her brother, husband and father – compares her grief to a “black ooze” that “revolts anyone who approaches”. She stays disengaged from Yalom, never meeting his eye, until one day he confesses he has just learned of the sudden death of his brother-in-law. Irene explodes: is he daring to compare his grief with hers? But from that moment on she expresses herself more sincerely, and begins to see that her anger protects her from further loss at the cost of keeping her alone. She learns to let her guard down, comes to weep in Yalom’s presence – and takes his hand when he offers her a tissue. In her final session, she asks Yalom if she can bring her new partner into the room to meet him.

Now 93, Irvin Yalom has captured better than anyone just how much drama and intrigue can be contained in the simple act of two people sitting together and talking in a room. In today’s world, that activity is becoming less and less common. Research suggests we spend less time in face-to-face interaction than any previous generation. More of us live alone, we belong to fewer associations, and we have fewer friends. Our growing isolation appears to be driven less by a lack of social opportunities than by choice: we seem to find the risk and friction of interpersonal contact increasingly unpalatable.

Our era’s psychological common sense – much of which can plausibly be traced to the individualist ethos of Carl Rogers and his school – may well be exacerbating such trends. We are supposed to guard our “boundaries”; identify “toxic” people and eliminate them from our lives; and “work on ourselves” to achieve “self-love” before we are fit to give and receive love with others. We hear much less about reciprocity, interdependence, repentance and forgiveness. These anti-social trends predated the internet, but there’s evidence they are being exacerbated by the era of smartphones and social media, leading to a mental health crisis among the generation raised on them. The average young person’s primary relationship – at least measured by the time they devote to it – is with the screen in their pocket. The results can be measured in growing rates of depression, anxiety, self-harm and suicidality.

Martin Buber, one of Yalom’s touchstones, warned against making people into tools. In the coming age of AI therapy, we will increasingly make tools into people, turning to screens to treat emotional problems associated with our screen-based lifestyles. Healthcare systems that currently provide therapy in the form of impersonal CBT (often over the phone or by text), and apps like BetterHelp that make therapists as replaceable as Uber drivers, will be able to do away with the inconvenience of humans altogether. Unlike the men and women they replace, AI therapists will be cheap, constantly available, never have off days – and will not suffer memory problems, grow old or die. They will offer us infinite positive regard, free from the problem of countervailing subjective feelings. Psychotherapists will once again become screens, expertly executing an algorithm, and we will talk to them while sitting alone in our rooms. In such a world, Irvin Yalom’s work will serve as a reminder of what we have lost.

[See also: David Rieff foretells the fate of woke]

Content from our partners
Better dementia diagnoses can lessen waiting list pressures
The government will fail to meet its goals without a solution to achieving good musculoskeletal health
Energy for a reset