Not too long ago, we were told social media would revolutionize human connection. It promised to bridge distances, foster deeper relationships, and build meaningful communities.
Fast forward to today, loneliness rates are skyrocketing, mental health crises are worsening, and authentic human connections have become rare in an increasingly digital world.
Now, the same big promises are being recycled but this time, with artificial intelligence (AI) therapy. 
The relentless narrative surrounding AI therapy is concerningly familiar.
Proponents claim it will democratize mental health care, providing universal, affordable access.
They argue that AI therapists are free from human bias and fatigue, available around the clock, and conveniently at your fingertips.
It sounds great, almost too great.
And indeed, it is.
Much like the early days of social media, the current hype around AI therapy overlooks fundamental truths about what actually makes therapy effective.
Therapy is not merely about receiving advice or solutions. It is deeply relational, rooted in empathy, shared experiences, and authentic human interactions.
It’s about a real person genuinely understanding another person’s pain because they’ve felt something similar themselves.
Reflecting on my own therapy sessions, a decade ago, I walked into my therapist’s office for the first time with a debilitating stutter.
Every conversation felt exhausting, each interaction fraught with anxiety and embarrassment. Through years of therapy, I experienced firsthand the irreplaceable power of human connection and empathy.
My therapist provided gentle, intentional guidance, patiently helping me trace my stutter back to specific childhood experiences.
Gradually, through trust, emotional understanding, and genuine human interaction, I was able to conquer my stutter, transforming my communication and life in ways an algorithm simply never could.
When my therapist shared a personal story that mirrored my experiences, the feeling was profound.
Knowing someone else had endured and overcome similar struggles helped me feel genuinely understood, validated, and less alone.
AI, in contrast, lacks the fundamental capacity to authentically relate to human experiences and emotions.
An AI repeating the same words as my therapist, devoid of genuine empathy or shared understanding, would feel hollow, sterile, even insulting.
Advocates of AI therapy often overlook the significant risks and lack of accountability that currently exist.
Who is responsible if an AI gives harmful advice? Who do we hold accountable when things go wrong?
The present state of AI lacks transparency, oversight, and ethical frameworks necessary for such delicate interactions.
Just as social media has evaded genuine accountability for its negative impacts, AI therapy platforms might easily follow suit, risking users’ mental health and safety.
Moreover, AI-driven interactions risk breeding more narcissism rather than reducing it.
A relationship with AI inherently centers around oneself, there’s no reciprocity, no genuine give-and-take.
Users may come away feeling superficially validated without ever being challenged, corrected, or genuinely seen.
This is the essence of narcissistic reinforcement: echo chambers of one’s own thoughts and biases.
Additionally, therapy isn’t simply about “feeling better.” It involves growth, which often comes from uncomfortable conversations and confronting truths.
Therapists challenge us, hold mirrors up to our blind spots, and help us navigate through discomfort.
An AI programmed primarily for user satisfaction and engagement is inherently incentivized to avoid tough conversations that might discourage continued use.
There’s also the troubling assumption that AI, by merely processing data and patterns, can genuinely comprehend complex human emotions.
Human emotion isn’t algorithmic. It’s messy, irrational, and deeply contextual.
The nuances of human experience, tone, body language, the subtleties of emotional responses remain inaccessible to AI, no matter how advanced its programming.
As we find ourselves at yet another technological crossroads, we must remember the failures and empty promises of social media.
Human connection cannot be replicated by lines of code. We shouldn’t be lured into replacing genuine relationships with AI-driven facsimiles.
AI may have its place as a supportive tool in mental health, offering supplemental resources, self-help techniques, and immediate assistance in emergencies.
But it must remain clearly defined as supplemental, not a replacement.
If we learned anything from the disappointing evolution of social media, let it be this: technology should support genuine human interactions, not supplant them.
We must resist making the same mistakes again by chasing the false promise of AI-driven connections.
The real work of healing is inherently human, and that is something technology simply cannot replicate.
