Interesting question. If you have a prior distribution for the space of possible functions, you can find the input that minimizes the entropy of the posterior d

Interesting question. If you have a prior distribution for the space of possible functions, you can find the input that minimizes the entropy of the posterior distribution (ie. makes it more like a delta and less like a uniform distribution)