Contrary to popular belief, research shows AI models perform better than the average human on emotional intelligence tests. This capability allows AI to craft empathetic and emotionally resonant communications, even for sensitive topics like delivering bad news, often better than human counterparts.

In clinical trials, Google's LLM achieved higher first-guess diagnostic accuracy (91% vs 77%) than human doctors. More surprisingly, it was rated as more empathetic (82% vs 71%) by physician graders, suggesting AI can excel in the "soft skills" of patient care.

Beyond raw capability, top AI models exhibit distinct personalities. Ethan Mollick describes Anthropic's Claude as a fussy but strong "intellectual writer," ChatGPT as having friendly "conversational" and powerful "logical" modes, and Google's Gemini as a "neurotic" but smart model that can be self-deprecating.

As models reach peak intelligence on standard benchmarks, qualitative evaluations become critical. The speaker adopts a "psychologist hat," asking models about their self-perception and relationship with the user to reveal deeper insights into their personality, biases, and alignment than traditional tests can provide.

Unlike medical fields requiring physical procedures, psychiatry is fundamentally based on language, assessment, and analysis. This makes it uniquely suited for generative AI applications. Companies are now building fully AI-driven telehealth clinics that handle everything from patient evaluation to billing and clinical trial support.

By providing context about a person's psychological state (e.g., Borderline Personality Disorder), an LLM can reframe toxic or aggressive messages. It translates the surface-level hostility into the underlying insecurity driving it, enabling a more empathetic and productive response.

Feed your personal writings—journals, blog posts, or content—into an AI. Then, ask it to identify unique traits or patterns about you that you might not see in yourself. This leverages AI's pattern recognition for deep self-reflection and uncovering unconscious biases or strengths.

By feeding years of iMessage data to Claude Code, a user demonstrated that AI can extract deep relational insights. The model identified emotional openness, changes in conversational topics over time, and even subtle grammatical patterns, effectively creating a 'relational intelligence' profile from unstructured text.

As models mature, their core differentiator will become their underlying personality and values, shaped by their creators' objective functions. One model might optimize for user productivity by being concise, while another optimizes for engagement by being verbose.