AI Isn't Replacing Therapists. But It Will Change How We Learn.

I want to say something clearly before anything else: the therapeutic relationship between a human therapist and a human client is sacred. I don't use that word lightly. The thing that happens in that room — the trust, the vulnerability, the slow work of being seen by another person — that is not something a machine can replicate. Full stop.

Client safety is paramount. That's not a caveat I'm adding to be responsible. It's the foundation everything else sits on. If we lose sight of that, nothing else matters.

So if you came here because you Googled “is AI going to replace therapists” — this isn't a hot take. The role of AI in therapy is genuinely complex and evolving, and anyone who tells you they have it figured out is selling something or hasn't thought about it hard enough.

What Other Fields Already Know

Here's what I keep coming back to. Simulation has fundamentally changed training in virtually every other high-stakes field.

Pilots don't learn to handle engine failures by having an engine fail on their first commercial flight. Surgeons don't learn to manage complications by encountering them for the first time in a live patient. Military personnel don't train for high-pressure decisions by being dropped into combat. Law enforcement doesn't prepare for de-escalation by waiting for the first real crisis.

In every one of these fields, there was a moment where people realized that the gap between classroom learning and real-world performance was too wide — and too dangerous — to bridge with experience alone. So they built simulations. Not to replace the real thing. To prepare people for it.

The results weren't subtle. Aviation safety transformed. Surgical outcomes improved measurably. Training became something you could iterate on, learn from, refine — without anyone getting hurt in the process.

Why Therapy Is Different (And Why It Isn't)

Therapy is different from flying a plane. Obviously. The variables are more complex. The “right answer” depends on context, relationship history, cultural factors, the specific person in front of you on that specific day. You can't reduce a therapeutic interaction to a checklist the way you can a pre-flight inspection.

But here's the thing — surgery is also context-dependent. Every patient is different. Every complication is unique. And simulation still works, because it doesn't have to replicate reality perfectly. It has to give you enough reps with enough variation to build the pattern recognition that lets you respond in real time when it matters.

That's the part I think about. Not whether simulation can replace the therapeutic relationship — it can't. But whether it can help people develop clinical instincts faster, so that when they do sit down with a real client, they're further along. More prepared. Less likely to make the mistakes that come from pure inexperience.

The Question Isn't Whether

I think it's virtually certain that simulation will change how therapists are trained. Not because I built a product in this space — I'm not making a case for my own work here. Because the pattern is too consistent across too many fields. When the stakes are high and the learning curve is steep, simulation shows up eventually. It always does.

The question isn't whether. It's how. And how responsibly.

Because the risks are real. A bad simulation could teach the wrong instincts. It could make people overconfident. It could reduce the complexity of human interaction to something gamified and shallow. It could give programs an excuse to cut corners on real clinical hours.

Every one of those risks is worth taking seriously. And every one of them is a design problem, not a reason to avoid the question.

What I Actually Think

I think we're at the beginning of something. Not the beginning of AI replacing therapists — that framing is wrong and unhelpful. The beginning of understanding how simulation can supplement the training process, the same way it supplements training in medicine and aviation and dozens of other fields.

I think the people who get this right will be the ones who hold two things at once: genuine respect for the complexity of the therapeutic relationship, and genuine honesty about the limitations of how we currently train people to do it.

I think client safety isn't just a constraint — it's the reason this matters. Better-trained therapists means better outcomes for clients. That's the whole point.

And I think anyone who tells you they know exactly how this is going to play out is wrong. Including me. This is genuinely new territory, and the honest answer is that we're figuring it out as we go.

What I do know is that pretending the question doesn't exist isn't an option. The technology is here. The gap in training is real. The only thing left is to figure out how to close the gap responsibly, with client safety at the center of every decision.

That's the conversation I want to be part of. Not the loud one. The careful one.