The invisible machinery behind therapy — and why much of it was never designed for the world we live in now.
Most mental health apps were built like tracking systems. The problem was that tracking behavior and changing it were different things — and the gap was where most products quietly fail.
AI is being hailed as the next frontier for early mental health detection, but its inherent 'sycophancy' creates a dangerous blind spot for complex risks like psychosis. We need a system designed to doubt its own intelligence.
Decades of psychotherapy research suggest the strongest predictor of success isn’t technique or technology. It’s the therapeutic alliance.
AI is already outperforming clinicians in trials. But when something goes wrong, who is accountable? From a global regulatory vacuum to the startups quietly building the safety stack, this is what a responsible AI mental health architecture actually looks like.
If you're building in mental health, digital health, or regulated care and don't have a product function yet, or have one that's not working, I'd love to hear about it.