This talk will be held online on Zoom, please register here to participate.

Abstract:

As autism diagnoses rise globally, healthcare systems face mounting waitlists and bottlenecks that delay access to care and support. Drawing on large-scale datasets (N > 660,000, PNAS 2018; N > 350,000, PNAS 2023), I will present evidence that autistic individuals, on average, show a heightened drive to systemize, experience difficulty with cognitive empathy (theory of mind), and yet maintain intact affective empathy. This cognitive profile emerges across daily life, from academic interests to responses to sound and music, and provides the foundation for a new AI-driven screening tool. Using people’s reactions to brief audio signals embedded in musical excerpts and machine-learning (N > 100,000), this tool predicts autism diagnoses and autistic traits with over 80 percent accuracy. This approach offers a rapid, non-invasive, and scalable pathway to timely identification and support for autistic individuals worldwide.