Discolab and AI for constitutional understanding

Discolab started from a simple premise: complex documents become more useful when people can read, compare, annotate, discuss, and rank them collectively.

During Chile’s constitutional process, the platform used AI-assisted search, summaries, comparison, and paragraph-level discussion to help citizens understand the proposed constitutional text in relation to previous texts and the current Constitution.

The platform connects three CRiSS-LAB interests. First, collective intelligence: comments, votes, and rankings reveal where groups agree, disagree, or need more information. Second, information accessibility: AI can reduce the cost of navigating complex legal documents when outputs remain linked to sources. Third, civic deliberation: public debate improves when participants can inspect evidence and compare interpretations.

Discolab also informs broader work on organizational learning, public consultation, and systems that help communities transform distributed knowledge into auditable collective understanding.

Coverage from UDD and La Tercera is available in News & Media.

Cristian Candia
Cristian Candia
Associate Professor, Data Science Institute, School of Engineering, Universidad del Desarrollo, Chile. Director of CRiSS-LAB.

Cristian Candia is a computational social scientist studying human and collective behavior with large-scale data, network science, experiments, and AI.