Title: Unveiling micro-signals of social tipping mechanisms in the climate discourse: Mechanistic insights using computational social sciences
Summary: This talk investigates how computational social sciences can identify “micro-signals”, subtle shifts in discourse, that signal social tipping points critical to climate action. Three studies highlight mechanisms linking societal dynamics to climate system resilience. First, using Ukraine’s Telegram data on climate-war discourse (2022–present), statistical changepoint analysis identified abrupt shifts (e.g., energy infrastructure attacks, dam destruction) tied to self-reinforcing feedback loops, information cascades, and crisis-driven reinforcement. These micro-signals act as early warnings of broader systemic shifts, aiding policymakers in crisis-driven transitions. Second, global Twitter/X data (2009–2021) revealed that conspiracy theories (e.g., “chemtrails”) dominate public resistance to solar radiation management (SRM), amplified by online toxicity. Positive sentiment emerged during governance discussions, underscoring transparency’s role in mitigating opposition. Third, a comparative analysis of 1.5M tweets on greenhouse gas removal (GGR) and SRM (2006–2021) showed GGR gaining traction through “natural” framing (e.g., afforestation), while SRM remained mired in conspiracy narratives. Together, these studies bridge social dynamics and climate systems, showing how computational methods trace tipping mechanisms, inform public engagement, and guide policy design for climate resilience. By decoding micro-signals, we advance predictive models of societal responses to climate interventions.
Ramit Debnath is a University Assistant Professor and Deputy Director of the Centre for Human-Inspired AI (CHIA). Ramit is an elected member of the Methods Advisory Group (MAG), Department of Work and Pension, UK Government. He leads the Cambridge Collective Intelligence and Design Group and climatRACES Lab, and has a visiting academic role at Caltech, where he co-leads the Climate and Social Intelligence Lab.
