Seminar: Improving Mental Health through Digital Text and Large Language Models

Language is the bedrock of human culture – it structures minds, shapes identity, and binds (or divides) societies. My work applies Natural Language Processing to explain and predict psychological states and behaviors, and uses large language models to enhance well-being and mental health. For example, analyzing Facebook can predict depression before clinical diagnosis. These methods scale to billions of posts to measure community-level health, well-being, and behaviors. Beyond measurement, I will discuss how LLMs can persuade individuals on policy positions and improve well-being. In mental health, in particular, LLMs may offer a generational opportunity to expand access to care, from training therapists to assisting with Cognitive Behavioral Therapy. Overall, I argue that the thoughtful combination of AI with psychological theory can meaningfully improve a wide range of societal outcomes.

Johannes Eichstaedt is an Assistant Professor of Psychology at Stanford and a Shriram faculty fellow at the Institute for Human-Centered A.I. He uses Natural Language Processing and LLMs to measure and improve mental health. In 2011, he co-founded the World Well-Being Project, now a consortium of psychology and computer science labs across four universities. Originally trained as a physicist, Johannes received his Ph.D. and clinical training at the University of Pennsylvania. Johannes was elected a rising star and emerging leader by three scientific organizations, has published 85+ articles, and has served as an expert for the United Nations and several governments.

Alternatively, join online:

Microsoft Teams

Meeting ID: 380 523 058 341
Passcode: f2Vv9rE7