We accept that any complex software, even well-designed and well-tested, can malfunction -- often spectacularly. What, then, can we expect from humans? Consider

We accept that any complex software, even well-designed and well-tested, can malfunction -- often spectacularly. What, then, can we expect from humans? Consider that virtually no effort went into the design; and the test framework is horrific, relying primarily on trial and error. The natural outcome is that any human "expert" is no more reliable than your average startup codebase. In fact, it's a wonder that human societies manage to organize themselves into anything resembling civilization.