Normal Accidents in the 21st Century: Artificial Intelligence, Social Media, and Mental Health

Laura Sills, University at Buffalo

Artificial Intelligence (AI) has become so embedded in our daily lives that its presence often goes unrecognized. The pervasiveness of AI is only compounded as industries continue to find new ways to incorporate this technology into their products, services, and existing systems. While some hail each use of AI as innovative, others raise concerns about the potential risks of AI. In this paper, I apply Charles Perrow’s theory of normal accidents to examine AI as a high-risk technology. I begin by illustrating how AI systems themselves demonstrate the necessary elements of interactive complexity and interdependence outlined by Perrow using the example of large language models (e.g., ChatGPT). I then turn to the integration of AI within social media, using Perrow’s framework to analyze social media algorithms and argue that the mental health crisis of Generation Z represents an unintended yet inevitable consequence of this high-risk technology—that is, a normal accident.

No extended abstract or paper available

 Presented in Session 159. Disparities in Recent Technological Health Advancements