Chang and collaborators confront an evidential field too large and dispersed for conventional synthesis. Their iconic idea is a global map of the co-impacts of natural climate solutions, assembled by using language models to screen more than two million articles and identify over 257,000 relevant studies. The theoretical contribution is to detach climate mitigation from carbon singularity: ecosystem interventions must be understood through biodiversity and human-well-being co-benefits, trade-offs and geographical asymmetries. Methodologically, machine learning becomes an instrument of evidence architecture, extracting locations, pathways, species and impact variables at a scale unattainable through manual review alone. The operation is cartographic redistribution of attention: it exposes mismatches between regions of high mitigation potential and regions of dense research, revealing that evidence abundance is itself unevenly produced. The wider bridge links conservation science, computational review, climate policy and postcolonial knowledge geography. The map therefore does more than organise literature; it makes visible how research capacity conditions which ecological futures can be justified, financed and governed.
URBANISM MEETS ART