A technical note about his note [1]. pg says, When that happens, it tends to happen fast, like a core going critical, but he probably means "like a core going supercritical", which is when a chain reaction goes on at an increasing rate (i.e. what you need for a nuclear explosion). Critical is an equilibrium (i.e the self-sustaining condition, as used for power generation).
Hmm. I could change it. But are you sure it wouldn't be hypercorrect to phrase it that way? Isn't going critical a kind of shorthand for going supercritical, because in practice any situation that produces the first is going to produce the second?
Edit: Actually I think criticality is the right metaphor. It's a change in state from what preceded it. That's the important part, not the resulting explosion. (Arguably there are in fact analogs to control rods in online discussions-- nesting depth, the fact that people get tired after a while, etc.)
Reaching criticality is a normal goal for the operators of a controlled nuclear reactor (nuclear power reactor unit; research reactor; etc.). A reactor that has reached criticality is usually one that is operating normally. For example: http://www.google.com/search?q=%22initial+criticality%22
The JRR-4, which reached initial criticality in 1965, had contenued to aperate safely until 1996 using uranium of a high concentration of 93% [...]
Unit 1 reached initial criticality on March 8, 1988 and went into commercial operation on August 25. Unit 2 reached initial criticality on March 12, ...
http://en.wikipedia.org/wiki/Critical_mass