All human knowledge relies on a gamble: “regularity exists“. Equivalently, only what is regular (e.g. a pattern), or what seems to be regular, has the right of entering our knowledge and scientific heritage. Think about the pervasiveness of the i.i.d. hypothesis in statistics and machine learning.
Note that regular does not mean necessarily something boring (a constant) or shallow or deterministic. We could find regularity in the behavior of a spring, as well as in the volatility of the stock market or in the way a complex dynamics evolves itself with time.
Nevertheless, humans consider that they know something only when they can fit it into a pattern, a model, a map. All the rest is unknown (or not yet known) and deserves labels like noise, error, uncertainty.
The regularity gamble 