It's essential: there's a reason we have physics, chemistry, biology, neurology, psychology, and sociology. (There are other similar chains of approximations.)

It's essential: there's a reason we have physics, chemistry, biology, neurology, psychology, and sociology. (There are other similar chains of approximations.) Each provides a tractably computable approximation to a scale that the preceding ones can't be worked through at, while also discovering emergent properties that manifest at that scale. There's no reason to think that our dominant paradigms in any of those fields are the only sensible way to approximate the underlying phenomena, and other (radically different) models might be able to express features present models struggle with.