There's a concept in performance psychology called deliberate practice.
It was popularized by Anders Ericsson1, the researcher who spent decades studying what actually separates experts from everyone else. His finding was counterintuitive: it's not raw talent, and it's not the number of hours you log. It's the quality of practice — structured, feedback-rich, just outside your comfort zone — that moves the needle!
Elite musicians practice specific passages, not full performances. Surgeons drill procedures on simulators before touching patients. Athletes break skills into components and work each one under controlled conditions.
And therapists?
For most of our field's history, therapists have practiced by seeing clients.
That's not nothing, experience matters enormously. But it's also not deliberate practice. Real clinical encounters don't pause so you can reflect on your response. They don't replay a moment you handled poorly so you can try again. They don't give you a resistant client on demand so you can practice therapeutic alliance repair when you're ready to work on it.
The result is a profession where learning is largely incidental, shaped by whatever walks through the door.
Simulation-based training has existed in medicine, aviation, and emergency response for decades. The evidence base is strong: practicing in controlled, high-fidelity environments improves performance in real ones. You build the pattern recognition, the emotional regulation, the reflexes, before the stakes are live.
Mental health training is catching up. And the tools are getting good!
AI-simulated client sessions can now replicate the texture of a real clinical encounter — the pauses, the deflections, the moments where a client tests whether you'll hold the space or fill it with noise. They're not perfect. But they're available at 11pm before a practicum placement, and they don't have real wellbeing on the line while you're figuring out your approach.
That's not a small thing.
Ericsson's research identified a few conditions that make practice truly DELIBERATE:
It has to be effortful. Comfortable repetition doesn't build skill, working at the edge of your current ability does.
It has to involve feedback. You need to know what you did, and what effect it had.
It has to be specific. Vague "practice" doesn't transfer as reliably as targeted work on a defined skill — open-ended questions, rupture repair, pacing, tolerance of silence.
AI simulation, at its best, can offer all three. The session is demanding. The transcript shows you what you said. You can choose the clinical scenario.
The best therapists we know (the ones decades into their careers who still light up when they talk about their work) share a quality that's hard to name precisely. They're still curious about what they might be missing. They seek out supervision not because they have to, but because they find it useful. They notice when they're on autopilot.
That orientation — toward learning, not just doing — is what keeps clinical work alive.
The tools for that orientation have never been more accessible. Whether it's AI simulation, structured peer consultation, or just a deliberate practice of reviewing your own sessions, the question isn't whether you have time.
The question is what you're optimizing for.
If you're curious what deliberate practice looks like in a clinical training context, we've been thinking about this a lot at TMind AI. We're not the whole answer, but we're part of a longer conversation about how the field trains the next generation of therapists.
If this resonates and you want to put it into practice, we've already mapped out what that can look like concretely — the goal-setting, the skill identification, the structure of building a deliberate practice framework into your clinical year. That's a good next read. 😃
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.