The miraculous universal distribution (Q1272403)

This describes a way of evaluating the randomness or complexity of data, of choosing between hypotheses, of assigning probabilities to hypotheses: by assigning probabilities to all hypotheses we satisfy Bayes' ideas, by choosing the most likely (the simplest) we use Ockham's razor. The probability comes from taking the negative of the complexity as an exponent: \(2^{-K}\), where \(K\) is some measure of the Kolmogorov complexity of the hypothesis. The purpose of these ideas is to draw conclusions from data in a reasonable way. Very readable and easy to understand at its own level.