Comment by alisonkisk

Why is frequentism bad because it only gives certainty for infinite samples, but complexity is good despite being non-computable?

It's two sides of the same coin -- computable uncertainty va non-computable certainty.