THERAPY SCHOOL UPDATE: I HAVE NOW BEEN CRITIQUED BY A ROBOT

A few people have asked what I am actually doing in graduate school, so here is the slightly long and winding answer.

I am taking four classes that look separate on paper but are beginning to feel like one enormous conversation:

• Systems of Family Therapy I
• Basic Techniques in the Psychotherapy Interview
• Systemic Approaches to Couples Therapy
• Cultural Competency

Systems is teaching me to stop looking at one person as “the problem” and start asking what relationships, histories, feedback loops, and larger systems are helping create or maintain the problem.

Couples Therapy has introduced us to the Gottman Method and Emotionally Focused Therapy. I completed and passed the Gottman Level 1 assessment, which means I can now barely hear sarcasm without seeing one of the Four Horsemen ride through the room. We are learning to look beneath what couples are fighting about and identify the emotional patterns that keep recreating the same conflict.

Basic Techniques is about the actual moment-to-moment work of therapy: listening, asking questions, reflecting feelings, tolerating silence, building trust, and learning when to speak and when to shut up. As it turns out, becoming a therapist involves a surprising amount of learning how not to talk.

Cultural Competency may be creating the most internal argument for me. We have been examining identity, privilege, oppression, power, cultural humility, and the ways therapists can unintentionally mistake their own worldview for objective normalcy. The assignments include a bias self-assessment, a diversity statement written for future clients, cultural reflection projects, and an ongoing journal about Jennifer Mullan’s book Decolonizing Therapy.

I have appreciated a great deal about Mullan’s argument. Therapy does not happen outside politics, economics, racism, sexism, ableism, class, religion, family history, or community. A person’s distress cannot always be understood—or treated—as an isolated malfunction occurring inside that person’s head. Sometimes the supposedly abnormal reaction makes perfect sense inside an unhealthy family, workplace, culture, or political system.

The book also challenges the idea that therapists are neutral observers. Therapists have power. They participate in systems that decide which emotions are acceptable, which behaviors are normal, who receives a diagnosis, and whose way of experiencing the world is treated as defective. That is something every therapist should take seriously.

My criticism is not that Mullan goes too far. In some places, I think she does not go far enough.

She often interprets these questions through her own experiences of otherness, which is completely understandable and often powerful. However, I keep thinking about people whose difference is harder to see or name. I think about autistic people who have learned to mask well enough to appear “normal,” including people who have never been diagnosed. They may experience years of small but pervasive social failures, misunderstandings, exclusion, shame, and exhaustion without anyone recognizing the actual difference underneath them.

Those experiences may not fit neatly into the cultural categories therapists already know to ask about. The danger is still the same: the therapist assumes that their own way of thinking, communicating, relating, or functioning is normal and then evaluates the client according to that standard.

That does not negate Mullan’s argument. It makes me want to push it one step further. If we are genuinely going to decolonize therapy, we have to question normalcy itself—including the normalcy inside therapists, researchers, diagnostic manuals, universities, and supposedly objective science.

All of that brings me to my first SimCare practice session.

SimCare is an AI program that plays the role of a client so we can practice conducting therapy sessions without accidentally damaging a real person while we figure out what we are doing. My simulated client was Reid Cohen. To be completely clear before anyone calls the ethics board, Reid is not a real client. Reid is a computer-generated person, although he managed to give me plenty to think about.

Reid had lost his job. He was drinking heavily, struggling to get out of bed, depending on his mother, feeling ashamed, and questioning whether he was a “real man.” He was also transgender, which meant that statement could have involved his gender identity, his ideas about masculinity and independence, his unemployment, his drinking, his relationship with his mother—or all of those things at once.

I thought I did a few things well.

At one point Reid said that he drank because he wanted the “noise” in his head to stop. I asked:

“What does the noise say to you?”

That question got underneath the drinking and produced one of the most important disclosures in the session. Reid said the noise told him that he was worthless, that he was a failure, and that everyone would be better off without him.

Later, when he said that his mother deserved a better child and saw him as a disappointment, I asked:

“Is that something you say, or something your mom says?”

I still like that question. It separated what his mother had actually communicated from the story Reid was telling himself about what she believed.

The program rated me highly in probing and clarification. It also credited me with using tentative language, avoiding “why” questions, respecting Reid’s gender identity, and not using stereotypes or obvious microaggressions.

Unfortunately, my therapeutic superpower and my therapeutic liability may be the exact same thing: I see six layers at once.

I saw the job loss, but I also saw depression, shame, alcohol use, purpose, creativity, family dynamics, masculinity, identity, isolation, and possible suicidality. Then I tried to talk about all of them in approximately 15 minutes.

I asked questions that contained two or three other questions hiding inside them. I would ask about what happened, how it felt, what caused it, and what it might mean—all before Reid had answered the first part. I was trying to dig deeper, but I was digging through too many layers too quickly.

Some of that may have exceeded what the AI was capable of following. But I cannot dismiss the criticism as a problem with the robot. A real client who is depressed, exhausted, drinking heavily, and unable to concentrate might also be overwhelmed by the way I was asking questions.

The point is not for me to stop seeing the deeper layers. The point is to let the client guide me into them.

I also talked about myself too much.

I told Reid that I had been fired, that I have bipolar disorder, and that I have spent much of my life feeling like I did not fit in. My intention was to establish genuine understanding and help him feel less alone.

The automated assessment labeled this “harmful self-disclosure.”

“Harmful” felt a little harsh at first, but I understand the criticism. I may have been trying to communicate empathy, but I was using Reid’s limited time to tell him about Francis. A disclosure can be sincere and still fail to serve the client.

That does not mean I have concluded that therapists should never reveal anything about themselves. It means I need to ask harder questions before doing it: Why am I sharing this? Whose need does it serve? Can I say it briefly? Will it deepen the client’s exploration, or will it make the client responsible for holding part of my story?

Another criticism involved Reid saying he did not feel like a “real man.”

I responded by affirming that I accepted him, but then I generalized the experience. I talked about how some people will not like us and how I have never felt that I fit in anywhere.

I meant to say, “You are not alone, and I do not see you as defective.”

What I should have asked was:

“When you say ‘real man,’ what does that mean to you?”

That would have let Reid tell me whether he was talking about being transgender, being unemployed, depending on his mother, feeling unable to control his drinking, or some combination of those things. I respected him as a person, but I did not fully treat him as the expert on the meaning of his own experience.

The most serious error happened immediately after one of my best questions.

When I asked Reid what the noise said, he told me that everyone would be better off without him.

I recognized that as possible suicidal thinking. I made the decision to build more trust before asking him directly about suicide. Then the clock started running out. I moved into other subjects and possible solutions, and I forgot to return to it.

That is not a small mistake.

I should have stopped and asked directly:

“When you say everyone would be better off without you, are you thinking about killing yourself?”

I had thought that asking too soon might damage trust. Looking back, taking his statement seriously and responding to it directly could itself have communicated trust, care, and safety. My job would not have been to “resolve” suicidal thinking in that moment. My job would have been to assess the risk and respond appropriately.

The other thing I noticed was how quickly my desire to help became an impulse to rescue.

By the end, I was suggesting staffing agencies, drawing exercises, employment possibilities, and future sessions with his mother. These were not terrible ideas. They were simply my ideas, offered before Reid had enough energy, clarity, or ownership to decide what he wanted to do.

That may be the biggest lesson connecting all four classes.

Systems tells me that the presenting problem is rarely the entire problem.

Couples Therapy tells me that the argument on the surface is often protecting something more vulnerable underneath.

Cultural Competency tells me that I cannot assume my understanding of health, success, identity, or normalcy is the client’s understanding.

Basic Techniques tells me that noticing all of that is not enough. I still have to slow down, listen, reflect, tolerate silence, ask one question at a time, and allow the client to lead me into their world.

I have joked about wanting a sign in my future office that says:

“I can’t fix your life, but you can.”

This exercise showed me that I have to learn how to live that statement in the therapy room. I cannot say I believe clients are capable of finding their own way and then spend the session racing ahead of them, identifying every hidden issue and handing them solutions.

I am not embarrassed by the simulation. I am excited by it.

The point of training is not to prove that I already know how to be a therapist. The point is to discover where my best instincts can become harmful when they are unrestrained.

I believe I respected Reid. I believe I empathized with him. I believe I saw important things beneath what he initially presented.

Now I have to learn how to see those things without rushing him, speak without taking over, share without making the session about me, and help without trying to become the person who fixes everything.

Apparently, therapy school is going to involve learning quite a lot about other people.

It is also going to involve learning an uncomfortable amount about myself.