Therapists are not Robots

Brace yourself for my bold and audacious statement, especially given that I am writing from one of the largest tech capitals of the world: there is a limit to what technology can do. *gasp* Before you come at me, let me preface: technology is invaluable. Humans have used technology since the beginning of time to advance society and facilitate the ease of complex tasks through the invention of tools like the compass, the clock, and the paper-mill. Digital technology is the next iteration of this advancement.

As technology develops, it does so at an exponential rate. One problem for our times to grapple with is what the potentially negative impacts of this rapid advancement will be. One possibility is that AI will replace some of our current workforce.

It has been estimated that 47% of US jobs may be in jeopardy by 2030 (Frey & Osborne, 2013). There has also been a recent expansion of AI bots like chat GPT. Some psychotherapy approaches are highly manualized and lend themselves to technological automation. Given all of this, it is logical to wonder if one day someone experiencing mental health concerns will be able to receive comparable outcomes from an AI bot as they would through a human therapeutic interaction.

As essential as technology is, it is just one way that humans grow. We also grow in vital ways through the humanities (arts, literature, philosophy) and in relation to others. Human relationships impact our physical health, our psychological development, and our sense of meaning in ways that non-human technology will ever be able to replicate.

In our hyper-productive, overachieving Western society, relationships can be cast aside as an afterthought resulting in social isolation or a subjective sense of loneliness. Approximately 1/3 of Americans aged 45 and older endorse feeling lonely (National Academies of Sciences, Engineering, and Medicine, 2020). One impact of this, as the CDC acknowledges, is that loneliness and social isolation put older adults at risk for serious medical problems including but not limited to dementia. These health impacts are even more serious for minority populations including immigrants and those in the LGBTQ community (CDC).

For many people, the Covid pandemic lockdown exasperated a sense of loneliness. Now that the pandemic is endemic, people who once flocked to remote work and telehealth for connivence are craving connection and prioritizing in person sessions once again. It is unlikely that individuals are motivated to seek in person sessions strictly for the physical health motives as described above. There is something innately transformative about being in the presence of a caring human being that compels some people to brave traffic and make it back into the therapy office.

Many theories of psychological change and personality development place relationships front and center. The theory of Object Relations at a basic level posits that we develop our sense of self by internalizing aspects of the relationships around us. Relatedly, the well-established Attachment Theory suggests that how we operate in the world and how we engage in our relationships is shaped by how important attachment figures show up (or don’t) for us. We are who we are, to a large degree, because of our relationships. Because human relationships have such a vital impact on psychological development, they also have an important role in healing.

Humans also possess skills that AI would have a hard time replicating. For instance, practitioners of Contextual Behavioral Science concern themselves with function (the underlying motive of a behavior) over form (what the behavior looks like). While AI might be able to detect the literal words that are being said and even perhaps the tone, it is through nuanced training and lived experience that humans are able to detect the deeper function of the words.

Beyond theory, the power of human connection is as real as our physical body. When a new life enters the world, skin to skin contact is encouraged as soon as possible to facilitate the release of oxytocin, the love and bonding hormone that is released through human touch. Then there are mirror neurons: neurons in the brains of some mammals (including humans) that fire when observing another engaged in an action or experiencing an emo>on.

Mirror neurons have implications for empathy and attachment. In the therapy room, my mirror neurons fire and I feel a deep sense of empathy when my parents experience suffering. Transformation can occur when my empathy helps my patients have empathy for themselves. How could this process ever be replicated by AI?

Relationships also impact us at a deep existential level allowing us to make meaning in this absurd world. Consider that when people are on their deathbeds, a primary reported regret is that they didn’t spend enough time with their loved ones. As Viktor Frankel, a psychiatrist that survived Auschwitz, observed, one of the reasons many who survived Auschwitz lived was for the hope of reconnecting with an important person in their life.

Relationships foster our will to live, and it is relationships that we have in view as we die. Relationships are at the core of human pain and human flourishing. Relationships drive us, shape us, correlate to our health, and contribute to us living more fully.

Given all of this, it follows that psychotherapy is effective to a large degree because of the therapists’ human qualities that foster the therapeutic relationship. It is well established in the research on therapeutic outcomes that “common factors” like therapist warmth, empathy, the therapist’s positive regard for the pa>ent, and the therapeutic relationship account for the most variance of change in therapy above and beyond what can be explained by any specific treatment model. AI may be able to replicate treatment models, but AI will never be able to replicate what really heals and transforms people. No mater how advanced technology becomes, unless it completely replaces humans, it will not replace psychotherapists.