In a recent postI mentioned four things therapy ignores his future career. Research on chatbot results, vulnerability of the 50 minute clock, what we really ask our customers and what competition affects pricing in every market it touches. This is my follow-up: what now?
First, some honesty about where I stand
I love my job and it has given me real meaning and purpose for 30 years. I’ve always said I’ll never retire, and I mean it. I believe that what the skilled therapists have to offer, AI can’t really replicate.
But I also think we overlook something that is rarely heard: therapy is not always a positive experience. I’ve done my own work over the years and I don’t look warmly or warmly on every therapist I’ve seen. gratitude. Some of these experiments were futile. One was worse. The idealized version of therapy that projects our careers, transformative relationships, and the healing alliance is real for many people, but it’s not universal, and I think pretending otherwise doesn’t serve our clients or our credibility.
A changed world
Some people always want to be recognized by another person, to sit across from someone who has their own human history and loss, who chose to do so for reasons that have nothing to do with an algorithm. It’s real and I can’t fathom it going away.
And yet it is also true that some people always prefer LLM. This is not a failure of treatment, but a reflection of a changed world. Especially young people have grown up with AI. For many of them, a chatbot doesn’t feel like a viable option. It feels familiar, private, and safe like sitting across from a stranger. We need to accommodate this reality, not deny it. But “some customers always prefer us” is not an effective strategy for meeting the future. It is comforting, and at the same time, comforting is not what this moment calls for.
Where the AI really falls short
Understanding where AI struggles is not only reassuring, but strategically important. Clinicians who develop true expertise in areas where AI is weaker are likely to be the best in the future.
A couple’s work is a good example. AI platforms are sycophants. They are designed to be compatible, affirmative, and avoid conflict in order to maintain their engagement. This is a serious responsibility in couples therapy, which often requires maintaining two conflicting realities at the same time, challenging both partners and enduring the discomfort of not taking sides. AI telling each partner what they want to hear is not a couple’s cure and can easily make things worse. The same is true of psychedelic-assisted therapy, which requires a depth of human presence and clinical judgment that no algorithm can provide.
Crisis intervention is another area of concern. Big language models can produce responses that sound reasonable but lack the common sense and real situational judgment that crisis work requires. They cannot read a room, hear a change in tone, calm a parent, or make the kind of split-second clinical judgment that a human therapist can when there is real danger.
Cultural sensitivity is also a significant challenge. These algorithms are trained on data that reflects the biases of whoever produced it. Clients from underrepresented communities may find that AI responses lose their cultural context. A therapist who has done culturally competent work brings important knowledge to this relationship that no current model can replicate.
What can we do?
Knowing where AI falls is directly where I believe therapists need to invest. Deepening skills in couples work, crisis intervention, injuryCultural competence and complex family systems are more than just good clinical practice. This is a good professional strategy.
AI literacy is also becoming more important. That means learning how to ask customers about their use of AI, how to assess whether it’s helping or hurting, and how to have informed conversations about what these tools can and can’t do. Clients are already using AI for mental health support, often without even realizing it. A therapist who knows how to bring this into their therapy room offers something truly valuable.
There are also real opportunities to work together with AI, essentially offering our clients the best of both AI and human therapy. Customized relaxation and mindfulness tools can be customized to the client and used between sessions, extending work without extending clinical hours. Therapy programs can provide support between sessions. Virtual reality offers new opportunities for exposure work and trauma processing with therapist guidance and experience integration. Therapists who learn to use these tools carefully offer a richer and more competitive model of care.
A note on ethics and responsibility
This is really new territory and the law hasn’t kept up with the technology. Different states have different rules regarding informed consent, therapeutic bots, and liability when AI is part of the treatment. If anything goes wrong, the responsibility remains with us.
Therefore, we must work carefully, document carefully and be present with the legal landscape of our states. We’re building a new path, and while that’s really exciting, it also means watching our backs. Ethical and legal mistakes can be so costly, and not all of our professional associations have provided us with the guidance we need to move forward. However, AI tools and hybrid therapy models are rapidly improving and cost a fraction of the cost of human therapy. This pressure will not go away. Therapists who manage it best, they make smart decisions rather than hoping the problem will resolve itself.
We are not alone in this
Of course, most, if not all, professions experience their own versions of these pressures, and many fear them. However, I believe that we are well equipped to meet this moment. We have spent our careers learning to sit with uncertainty, tolerate not knowing, and help others. fear to something more workable. That is exactly what this moment demands of us.
Let’s stop saying “AI won’t replace your job” and get up to it. Only in this case we can become nannies for the future instead of the traces of another time.
