Suicide Prevention and AI

In this course, Dr. Jonathan Singer, Ph.D., LCSW, Loyola University of Chicago School of Social Work, introduces the notion of AI-assisted prediction of suicide risk by citing research that showed traditional approaches to prediction have been no better than chance. Contextualising his talk with an outline of what tasks AI can do, Singer distinguishes between data sourced from medical files, social media entries, and geolocation data, noting that when these are integrated, the lethality gap present in medical data alone may be eliminated. Dr. Singer explains the difference between the various types of natural language processing, and how these can be used in a therapeutic context. The potentials of digital therapeutics and teletherapy are also addressed, along with ways that mental health professionals can use AI. The lecture balances the positive possibilities with a summary of numerous limitations to keep in mind as we consider using AI to prevent suicide.