"High reliability is not the absence of errors, but the presence of capabilities that detect and contain them.
Weick is not describing a world without errors. He is describing an organization that stays functional when errors happen, because it detects them early and contains them before they cascade. Errors happen; the difference is whether the system notices early.
That capability is built out of choices that can feel like overhead in calm weeks: tight feedback loops, small blast radius, clear escalation paths, and defaults that favor containment over heroics. When you cut these first, you are spending reliability without noticing.
The standard to defend is not no incidents. It is detect and contain. Make the exception path explicit, because hidden exceptions are the fastest way to lose the capability you thought you had. If exceptions happen by vibes, you are building fragility.