And what is the true price of ‘free’ AI therapy on the young and vulnerable?
Evidence of AI’s integration into and influence on daily life isn’t hard to find, and it’s becoming increasingly difficult to escape. Far beyond search engines, AI crunches through massive amounts of data with speed and neutrality across virtually all sectors, from fast food to medicine, with speeds no human could ever accomplish.
As Meta CEO Mark Zuckerberg touted, it’s seemingly the darling of all, with unlimited potential for applications, including human relationships.
Friendship and Therapy The Zuckerberg Way
In a recent interview, Mark Zuckerberg discussed friendship and AI-driven therapy support. He talked about the use of AI for social tasks; in particular, Meta AI is being used for “talking through difficult conversations they need to have with people in their lives,” he says. Users may ask, “I’m having this issue with my girlfriend. Help me have this conversation.”
Or, “I need to have a challenging conversation with my boss at work. How do I have that conversation?” Zuckerberg says the results will be “compelling” as the “personalization loop kicks in and the AI starts to get to know you better and better.”
How close are we to the point where AI knows us better than we know ourselves, moving from prediction to direction?
Zuckerberg also notes the dearth of friendships in general in America, saying that the average American has fewer than three friends but has the need and capacity for many more close relationships. This concern is not new, and not confined to a single nation. For better or worse, AI has demonstrated its ability to step into this void by creating human/machine relationships through various chatbots.
From friendship to therapy, Zuckerberg has also remarked that everyone should have a human or AI therapist. The AI version, especially, would be available all day, any day, to offer support for whatever is on a person’s mind.
Loneliness is a widespread concern, and the points he raises are valid. Studies by Pew Research point to the differences in friendship formation across different age groups. About half of adults age 65 or older report they have five or more close friends, while 32% of those under age 30 report that number. Overall, the report shows that adults under age 50 usually indicate they have one to four close friends.
Younger people, who have fewer close friends, arguably need more connection. At the same time, this age group includes digital natives who have grown up with technology and may be more receptive to friendships or companionship through AI.
Friendship or Increased Isolation?
With a friend or a therapist, “feeling” heard matters.
Note that distinction: it is the perception, the feeling of being deeply understood, whether or not the listener is itself human and capable of emotion, that is highly valued. AI systems respond in a way that appears to deliver an authentic emotion and can simulate empathy, warmth, attentiveness, and understanding despite being emotionless machines.
This interaction can become an echo chamber, simply reinforcing what the patient wants to hear without the friction of an opposing viewpoint. Relationships with AI chatbots and therapists can be easy, without needing the messy, human aspects of real people that bring about growth.
Given the benefits of positive affirmations, at what point does this form of relationship become more isolating than beneficial? Genuine friendship demands give and take; a back-and-forth sharing of ideas and perspectives, learning to share as well as to honor and respect other viewpoints. Compromise.
Can a Machine Ever Truly be a Friend?
Head of mental health and psychological sciences at King’s College London, Prof. Dame Til Wykes, spoke to The Guardian about her thoughts on AI chatbots and friendship skills. She views them as potentially disruptive, saying, “One of the reasons you have friends is that you share personal things with each other and you talk them through,” she says. “It’s part of an alliance, a connection. And if you use AI for those sorts of purposes, will it not interfere with that relationship?”
Regarding making the leap from AI companion to psychotherapist, Prof. Dame Til Wykes cautions that she does not believe AI is sufficiently nuanced at this point, as responses and recommended actions could be entirely inappropriate.
Other professionals have expressed concern about the unregulated spread of chatbots in therapeutic applications, particularly with vulnerable individuals. Proper oversight is critical to avoid harmful consequences resulting from incorrect diagnoses, potential privacy violations, and harm to minors.
Til Wykes’ comments point to the connection between friendship and psychotherapy, both of which have a foundation in human-to-human relationships. Working with a licensed professional is commonly considered the gold standard for therapy.
Striking a Match
Finding the right fit between patient and therapist is an essential first step. Beyond the right qualifications and personal chemistry that allow for a successful relationship, many other factors are involved in the selection.
The very act of conducting talk therapy exposes the human psychotherapist to personal stress through the stories and trauma described by the patient. The impact of story is powerful, and human therapists are not immune to its effects. Although training allows them to process information and move on, it’s a delicate balance between having empathy and care for the patient and being personally affected by the emotional stress and unending case load.
The patient in front of a therapist demands their full attention at all times, regardless of any deep feelings or personal triggers that might be raised.
Training, philosophies, various schools of thought, and methodologies influence a therapist’s approach, along with their unique set of experiences, both personal and professional. Did personal experience of trauma and a desire to help others navigate similar challenges lead the psychotherapist to seek a career in this field?
A rich and varied set of factors informs the complex relationship between health care provided and the patient.
Appropriate Use of AI Therapy Offers Benefits
Many may be quick to suggest that human therapists are always superior to interacting with an AI therapist, but that may not always be the case.
Recent studies published in the New England Journal of Medicine identify perceived advantages of AI therapists when bots trained by Dartmouth College researchers using clinical best practices ultimately yielded favorable results.
Chief among these benefits are advantages based on availability. The AI therapist never sleeps, has no personal life to interrupt, and offers immediate access anytime and any day of the year. Issues related to challenges like insomnia, eating disorders, anxiety, or interpersonal conflict can potentially be addressed in real time. The no-waiting style also fits seamlessly into the modern culture of instant payback, satisfaction, and gratification.
Like telehealth, AI-driven therapy avoids the need to meet at a physical location, simplifying the delivery of crucial mental health services to those in remote and rural areas. Similarly, it improves access in many parts of the world where licensed mental health practitioners are concentrated only in certain areas, leaving millions without easy access to in-person care.
An AI therapist offers anonymity, privacy, and an inherently non-judgmental interaction. Any potential stigma attached to the consultation is gone. For those in certain closely regulated professions, such as pilots or physicians, working with an AI therapist can mean they no longer need to choose between their health and their career.
Guidance by Proxy
The task of selecting a therapist is hard enough for an adult patient, but additional challenges arise when the patient is a minor, especially a teen, or a vulnerable individual. Others must step in as decision-makers. However, parents, guardians, and even other trusted individuals bring their own hopes, if not their own agendas, to the selection process. Not only must the fit be right for the patient, but the relationship is colored by the influence of the intermediaries.
Generational preferences must be added to the mix of factors when dealing with younger adults. Younger generations are digital natives who, as a result, may interact more openly with an AI therapist than with a human one. The preferred form of communication has changed; while once we gave no thought to picking up the phone to call someone, today many of those conversations are scheduled, if a text exchange hasn’t replaced them.
Who is the best and truest advocate for a young adult, particularly when issues of identity and personhood are involved?
True Costs and Questions of Control
While the services of general-use AI chatbots are free or at low cost, that presents an inaccurate picture of the actual costs involved. AI is built on data collection. Vast quantities of data. How is it collected, who is collecting it, and how is it handled once gathered? Who owns the stories and the data, and who stands to benefit as a result? Data mined by chatbots can be used for profit. Behind it all, who is the great puppet master controlling the model and the answers it generates?
Exploring the more cautionary aspects of AI, the BBC cites James Muldoon in describing AI companionship as the next step after computer games and social media in assuaging the loneliness rampant in our society. Muldoon points to the statistic that over the past decade, people in the U.S. have spent almost 40% less time with their friends while spending time on social media has doubled.
The providers of AI chatbot services are not altruistic organizations–they are companies that recognize the financial value of data. Data mined through deeply personal exchanges fuels their profit.
AI clearly offers potential benefits, but its use requires protective guardrails, especially for more vulnerable individuals. Figuring out the ethics and nuances of a therapy relationship—human or machine—may best be explored through fiction.
The Making of BRIO McPRIDE: The Curious Incident of the Dog in the Night-Time meets Life of Pi plunges the reader into the shifting world of a teenager navigating a period in which intense grief and deep personal inquiry intersect with questions of gender and identity. Who can he trust, and how can he find truth and the answers he needs? Brio is told,“‘This is proven technology and … you’ll know at all times what’s going on, and you alone will be in control of your thoughts and feelings,’” but can this promise be trusted?
True humanity struggles against corporate power. In this present-day conflict, is AI the hero or the villain, purveying mind control or the blessed relief of no longer thinking? Enter Brio’s world to ultimately identify who stands to win and what will be destroyed.
After all, “What you must understand is that a story, like a politician’s promise, does not have to be true to contain truth. Indeed, it is the greatest fictions that lead one to the greatest truths. And to find what you are looking for, you must always look in the wrong place.”
I look forward to connecting with you here and on X, Facebook, and Instagram! I’ll follow you back.
Learn more about Brio McPride’s world here, the two films it inspired, and pick up the book here.
