They’re not, but they’re spectacularly good at spotting dumb errors that humans are spectacularly bad at spotting.
Me, I still haven’t got my head around how to prove something on a computer, but the principle is sound. Theoretically one could build the proof as some form of literate program.
A computer just does symbolic manipulation according to a list of rules (axioms) and the human/programmer specifies which sequence of symbolic manipulations to apply and then the computer simply states whether or not the specified manipulations transform the theorem in to the truth symbol.
Me, I still haven’t got my head around how to prove something on a computer, but the principle is sound. Theoretically one could build the proof as some form of literate program.