Remember, if someone tells you that the chances of something happening are "next to nil" or "almost impossible" or even "highly unlikely," then the chances are still non-zero.
Next to nil isn't zero.
This axiom is clue that something is at risk and you should pay attention to it. It's not statistics that we should worry about it. It's our assumptions.
From: Lex Fridman's John Carmack Interview
It's not the writing of the program initially, it's the whole lifespan of the program.
No software project worth putting into production is ever truly static. Even a hypothetically bug-free application upon initial release has to change as it gets used in the real world. Users provide feedback so the UX has to change; the product owner or client of the application wants to tweak something; a new platform comes out for it to be deployed to; the app is so popular that it needs to be scaled up — any number of circumstances will force a change in software at some point during its lifetime.