Mathematical Patterns vs. Human Presence

Is talking to an AI system the same as talking to a therapist?

In the first episode of AI and Psychotherapy, Alejandro and Ishita tackle the growing trend of using large language models (LLMs) for mental health support. Ishita breaks down what psychotherapy actually is and why it goes far beyond just feeling better, while Alejandro explains how these models work under the hood and why their design to be agreeable can become dangerous in vulnerable moments.

They discuss the sycophancy problem, where AI validates instead of challenges, the privacy risks of sharing your deepest struggles with a cloud-based system, and a compelling study showing that AI and human therapy combined outperform either alone. If therapy is too expensive for many people, is an imperfect AI tool better than nothing at all?

An honest conversation between two people driven by a shared belief that understanding both sides is the only way forward.