Thought experiment: option 1 is chosen, and the world proceeds to spend the the next 50-100 years obsessing over whose WMDs, exactly, have any nonzero chances of ever, ever misfiring.

Humanity embarks on a top-down superoptimization campaign to perfect technology to the point that this question can be answered with ever-increasingly-higher levels of detail.

Follow-on: the question of whether the ensuing yak shaving causes us to realize the WMDs are unnecessary, or whether military-driven paperclip maximization accidentally wipes out humanity.