It is absolutely fascinating how wrong this approach is. Every single issue that comes up in on-call should be evaluated under the lens of “does fixing this ab

It is absolutely fascinating how wrong this approach is. Every single issue that comes up in on-call should be evaluated under the lens of “does fixing this absolutely require human judgement, or can it be automated, ideally by fixing the code in the main system. If it does require human judgement, are there ways to redesign so that is no longer true?”