The bottleneck in mental health isn’t treatment.
It’s trained people.
We know what works. There aren’t enough practitioners who can do it well — and the way we train them hasn’t fundamentally changed in decades.
CBTLAB is a training environment where practitioners build clinical skill through simulation and deliberate practice, with instant, criterion-based feedback.
There is no flight simulator for this profession.
Therapists largely do learn on real clients — with supervision that is scarce, expensive, and delayed by days or weeks. A trainee might practice a specific skill a handful of times before using it with someone who needs it.
Pilots don’t learn by crashing planes. Surgeons don’t learn by operating unsupervised.
That’s the gap we’re building for.
Three positions, one lesson.
Each lesson trains a single competency, and the learner occupies three positions at once.
- 1 They watch it done. A worked example, narrated step by step.
- 2 They supply the material. A real, ordinary moment from their own week — which is what makes the practice real rather than hypothetical.
- 3 They do the work. They run the actual clinical moves on that material, and get told specifically which parts of the skill landed.
The third position is the point. You can explain a technique from a textbook. You can only perform one after doing it.
The pedagogy is chosen, not improvised.
Learners practise the whole skill from the first lesson, with support that fades as they improve. Every skill is demonstrated before it’s attempted. Agreement never counts as progress — the system records learning only when the learner produces something in their own words. Feedback names the specific criterion met, quoting the learner’s own words as evidence.
What CBTLAB is not.
Not therapy. Not a replacement for clinical supervision. Not a diagnostic tool.
It’s skills training — established educational science, made interactive by AI.