It’s because the non-broken alarms, which also start out loud, get quieter throughout your life as they calibrate themselves, and as one’s habits fix the situations that make them correctly go off. So given a random initial distribution of loudness, eventually the alarm that’s loudest on average will probably be a broken one.
There are two distinct questions. First, are there quiet broken alarms. I propose that there are and your theory should agree—the person starts off with a broken alarm, modifies their behavior to the extreme, and the alarm stops coming up. Your text seemed to imply that it can’t stop, but I don’t think I understood why not.
Second, are there loud correct alarms. It easy to imagine a case where there would be. For example, an alcoholic might worry that his drinking is hurting his family. Or an introvert could worry that they’re too boring to hang out with. Your suggestion that habits are easy to change is weird.
This is related to simultaneous overconfidence and underconfidence. The loudest alarms are very probably too loud. The most needed alarms are very probably too quiet. I don’t think removing the habits-easy-to-change assumption really impacts this. Yeah, there are cases when your worst problem will stick out like a sore thumb because it is very bad. But, if there are a large number of possible problems and there’s some noise in alarm calibration, it’s unlikely.
I agree there are broken alarms that are quiet (including those that are broken in the direction of failing to go off, which leads to a blind spot of obliviousness!), and that there are people stuck in situations where there is a correct loud alarm that happens most of the time.
I said that habits are easier to change than alarms, not that they’re easy in an absolute sense.
It’s because the non-broken alarms, which also start out loud, get quieter throughout your life as they calibrate themselves, and as one’s habits fix the situations that make them correctly go off. So given a random initial distribution of loudness, eventually the alarm that’s loudest on average will probably be a broken one.
There are two distinct questions. First, are there quiet broken alarms. I propose that there are and your theory should agree—the person starts off with a broken alarm, modifies their behavior to the extreme, and the alarm stops coming up. Your text seemed to imply that it can’t stop, but I don’t think I understood why not.
Second, are there loud correct alarms. It easy to imagine a case where there would be. For example, an alcoholic might worry that his drinking is hurting his family. Or an introvert could worry that they’re too boring to hang out with. Your suggestion that habits are easy to change is weird.
This is related to simultaneous overconfidence and underconfidence. The loudest alarms are very probably too loud. The most needed alarms are very probably too quiet. I don’t think removing the habits-easy-to-change assumption really impacts this. Yeah, there are cases when your worst problem will stick out like a sore thumb because it is very bad. But, if there are a large number of possible problems and there’s some noise in alarm calibration, it’s unlikely.
I agree there are broken alarms that are quiet (including those that are broken in the direction of failing to go off, which leads to a blind spot of obliviousness!), and that there are people stuck in situations where there is a correct loud alarm that happens most of the time.
I said that habits are easier to change than alarms, not that they’re easy in an absolute sense.