Maybe there’s a bit of terminology confusion here; if a military conflict /doesn’t/ affect me personally, it seems unlikely to be a ‘significant’ one. (Some historical ways a military conflict could affect me personally: Victory Gardens, the Order of the White Feather, the Fenian Raids, even less oversight and accountability for civilian police whose actions would otherwise end up in the subreddit “Bad Cop, No Donut”.)
I’m thinking of scenarios such as ‘It turns out China put secret backdoors into all sorts of hardware chips, and in a fit of self-righteous pique (which they think will play well to their red-state base), the war-monger side of the American Congress doesn’t see any downsides to making a demand that everyone in the world shut down their supposedly Chinese-controlled hardware under threat that if they don’t, they’ll send the American military to shut it down’. As far as I can tell, several versions of just this one particular scenario don’t obviously break the sociological law of every political actor having to act in what they perceive to be their own self-interest.
However, I no longer trust my sense of calibration for the odds of large-scale politics, given that I was willing to go along with the predicted odds of 88% for Hillary winning the election, and didn’t update nearly as much as I should have by the time of the election itself. And said lack of calibration puts a sharp limit on how rationally I can act as I decide how much effort to put into preparing for the more unpleasant scenarios.
You sound paranoid. Even if there is significant military conflict, it won’t affect you personally.
Maybe there’s a bit of terminology confusion here; if a military conflict /doesn’t/ affect me personally, it seems unlikely to be a ‘significant’ one. (Some historical ways a military conflict could affect me personally: Victory Gardens, the Order of the White Feather, the Fenian Raids, even less oversight and accountability for civilian police whose actions would otherwise end up in the subreddit “Bad Cop, No Donut”.)
I’m thinking of scenarios such as ‘It turns out China put secret backdoors into all sorts of hardware chips, and in a fit of self-righteous pique (which they think will play well to their red-state base), the war-monger side of the American Congress doesn’t see any downsides to making a demand that everyone in the world shut down their supposedly Chinese-controlled hardware under threat that if they don’t, they’ll send the American military to shut it down’. As far as I can tell, several versions of just this one particular scenario don’t obviously break the sociological law of every political actor having to act in what they perceive to be their own self-interest.
However, I no longer trust my sense of calibration for the odds of large-scale politics, given that I was willing to go along with the predicted odds of 88% for Hillary winning the election, and didn’t update nearly as much as I should have by the time of the election itself. And said lack of calibration puts a sharp limit on how rationally I can act as I decide how much effort to put into preparing for the more unpleasant scenarios.