conservatism

There are two categories of approaches for apprehending the world: rational and empirical. Rationalists have a cognitive bias towards thinking in terms of models, systems, and logical consequences. As a result, they tend to develop models of reality based on a limited set of direct sense data and then try to impose that highly developed, internally consistent model onto their subjective experience of the world, filing down the edges where things don't fit. Rational models tend to be brittle, demonstrating a preference for rejecting contradictory information rather than assimilating it for fear of breaking the model altogether.

Empiricists approach the entire project of making sense of subjective experience from a different starting point. Firstly, they take in more sense data before attempting to build an explanation. Second, the model they end up building often contains internal contradictions, exceptions, and noise. The model empiricists build is inherently more complex, more opaque, and amenable to incremental adjustment. It also often demonstrates more explanatory power, but only after significant iteration.

When rationalists encounter empirical models, they are frequently thrown off by their messiness. Their initial reaction is likely to reject the model as nonsensical, dumb, or both. It is the sort of thing one experiences when trying to learn English as a second language, since English is notoriously irregular as human languages go.

In the realm of politics, what a lot of liberals and left-leaning types misapprehend about conservatism is that conservatism is not an ideology, it is a disposition. It is a fundamentally different way of building a model for how the world works that is often contradictory, messy, and opaque - much like humans themselves.

Neural networks only work when they are trained on enormous amounts of data, at which point they far surpass anything an algorithm can do.