> More seriously, it’s reasonably well-known among mathematicians that published math papers are full of errors.[1]
I think the author's point is that actually the distinction is less clear: many math papers can't be (easily) replicated, yet people in math aren't too worried.