[ratmice]

My only complaint with the article is that it doesn't seem to mention that digitized proofs can contain gaps but that those gaps must be explicit like in lean the `sorry` function, or axioms.