We will cover three foundations for prediction and classification. generated signals, categories, and dimensional reductions. These are used to model humans, other fauna, standard statistical predictions and AI. We will then examine a variety of theorems including Condorcet Jury Theorem, The Diversity Prediction Theorem,
The Bias-Variance Decomposition Theorem, The Category Prediction Theorem, The Two-Population Theorem, and the recent Hong Page Diversity-Accuracy Classification Theorem. We will conclude with two evolutionary models that link institutional and network structures to diversity levels, and therefore, to collective accuracy.