WeDecide is building civic tech for responsive government.
How do we measure whether an institution is actually learning?
Traditional approaches often rely on retrospective measures such as satisfaction, consultation participation, or sentiment. We are developing a different approach: measuring the relationship between institutional expectations, incoming evidence, actions, and verified outcomes over time.
We are seeking a Civic AI Researcher to help develop the probabilistic and computational foundations of this system.
What You’ll Do
- Model institutional prediction error: Develop methods for comparing prior institutional expectations with observed outcomes and determining how those models should update over time.
- Develop evidence-weighting models: Investigate how institutions respond to conflicting or disconfirming evidence, including changes in response latency, categorisation, escalation, and evidence weighting.
- Separate belief, action, and outcome: Build methodologies that distinguish what an institution predicts, what it records, what it actually does, and what happens afterward.
- Develop our probabilistic civic-learning layer: Apply techniques from Active Inference, Bayesian modelling, computational social science, causal inference, and related disciplines to real-world governance systems.
- Advance Symbiotic Intelligence: Work closely with our Chief Science Officer, engineering team, and external research partners to translate an emerging scientific framework into testable models and deployable product infrastructure.
- Bridge research and deployment: Help ensure our models remain scientifically rigorous, interpretable, auditable, and practical enough to operate inside real institutions.
Who You Are
You have a strong academic or applied background in one or more of the following:
- Computational social science
- Bayesian modelling
- Complex adaptive systems
- Collective intelligence
- Mechanism design
- Algorithmic governance
- Causal inference
- AI and public institutions
You are interested in applying rigorous theoretical frameworks to real institutional systems rather than leaving them solely in academic literature. You care about interpretability, measurement, falsifiability, and real-world outcomes. And you are excited by the opportunity to help build a new scientific and technological framework for understanding how institutions learn. If that sounds like the kind of problem you want to spend your time on, we would love to hear from you.