1. Conversations with science: helping AI to unlock the biology of mental health for public understanding (2026–2027)
We want to help AI-powered conversational search tools (like ChatGPT, Claude, open source Large Language Models, etc) be able provide accurate answers specifically for mental health science questions, to help people instantly explore the biology of psychiatry without needing to understand complex data systems. But an AI is only as smart as the questions we teach it to answer.
It ensures the AI answers real, human questions (e.g., “How does stress alter inflammation in depression?”) rather than just clinical jargon.
It helps map out how advanced biological data can be surfaced quickly during patient care or discovery sessions.
You will help us decide what people actually want to ask about the biology of mental health, test early AI prototypes, and define what a “good, trustworthy, and understandable answer” looks like.