Join us online! Grab your lunch and tune in for a session at the intersection of AI and evidence-based policymaking.
Presentation title: Large language models can accelerate evidence-based policymaking
Policymakers face an unprecedented challenge where evidence-based policymaking becomes increasingly impossible because of the intrinsic challenges associated with identifying and synthesising evidence within a marshland of noise. One solution relies on the use of expert groups and consensus methodologies, exploiting the information processing and synthesis contained within an expert opinion. Unfortunately, these processes are resource-intensive, contain bottlenecks, and risk sampling bias because costs escalate when more expert opinions are elicited.
Large language models (LLMs) offer new opportunities to systematically integrate diverse expert perspectives at scale. We developed Symphonia, an open-source platform, to test the hypothesis that AI can efficiently and robustly automate iterative consensus methodologies. We demonstrate Symphonia’s application using a nationally significant, policy-relevant case study. Experts evaluated the validated platform’s effectiveness, highlighting its potential to accelerate structured synthesis without compromising rigour.
Our findings demonstrate that LLMs enable faster, less resource-intensive, more transparent, and more scalable academic consensus synthesis.
Guest Speaker:
Professor Mark Mon-Williams
University Director of the Wolfson Centre for Applied Health Research, Impact Director for Medicine and Health at Leeds, Executive Member of the Born in Bradford (BiB) longitudinal birth cohort programme, Child Health Outcome Research At Leeds (CHORAL), and the N8’s Child of The North partnership. Fellow of the Royal College of Paediatrics and Child Health and leads the BiB Centre of Applied Education Research and NHS Yorkshire Applied Research Collaboration ‘Data Analytics, AI and Visualisation’ group. He is Deputy Chair of the Department for Education’s Science Advisory Council and Chair of the Deputy Prime Minister’s ‘preventative analytics’ taskforce.