On: Why is writing mathematical proofs more fault-proo...

[setra]

These days programming languages actually are used for expressing proofs. They can automatically check them to. For example the Coq theorem proving language.

https://en.wikipedia.org/wiki/Coq

Some would argue that writing code IS more fault proof than writing proofs the traditional way.

[contravariant → setra]

I would still trust a mathematical argument I can understand over the output of some automated theorem prover.

Of course these aren't mutually exclusive, provided the theorem prover is simple enough to understand.