If axioms are mutually contradictory, that doesn't actually prevent the concept of computing them and attaching them to logical chains.

The core problem with computing them is that no machine can decide whether the axioms are contradictory in the first place. In most cases, even humans won't 100% agree.