3 ms·
But a "computer running a program" is not an interesting proof of anything. The proof depends on how the program is written. For example, a program written in a
by pron 10y ago
But a "computer running a program" is not an interesting proof of anything. The proof depends on how the program is written. For example, a program written in an untyped language encodes only a very trivial proof by the C-H correspondence (of `tt ⇒ tt` or something like that) even though it performs the same computation of a program written in another formalism that expresses a much more interesting proof. In other words, the proof does not lie in the dynamical process of computation, but in properties of the language used to express the computation. This is a much less universal claim.