Re:Crashing, Burning Planes Don’t Stop Passengers from Grabb...
by piotrkaminski on Thursday August 04, 2016 @05:00PM
(#12228446)
> Meanwhile the makers of the alarms have an incentive to be overly sensitive because a false positive is much less dangerous than a false negative.
I wonder whether this is actually statistically true, if false positives cause true positives to be ignored by some non-trivial proportion of the population? I think it's just as likely this is legal CYA from the alarm makers -- they can only get sued for false negatives, after all.
Re:Crashing, Burning Planes Don’t Stop Passengers from Grabb...
by saulrh on Thursday August 04, 2016 @05:03PM
(#12228468)
To them a false negative is worse than a false positive, for precisely the reason you describe: it simply doesn't matter how many false positives you have, but the very first false negative destroys your company. Their optimization function is not fire safety. It's making money, which is only mostly associated with fire safety.
Re:Crashing, Burning Planes Don’t Stop Passengers from Grabb...
by piotrkaminski on Thursday August 04, 2016 @05:15PM
(#12228539)
Right, so I wonder how we could change the incentives so their optimization function is a little more socially responsible? Perhaps a law that allows civil lawsuits for false positive alarms would help, but I'd be afraid of unexpected side-effects.
Re:Crashing, Burning Planes Don’t Stop Passengers from Grabb...
by colanderman on Thursday August 04, 2016 @08:59PM
(#12229471)
It's the topic of one of Aesop's fables, so it's at least a widely observed phenomenon: https://en.wikipedia.org/wiki/The_Boy_Who_Cried_Wolf
A similar problem exists with car alarms. What percentage of car alarms going off are in response to a legitimate threat of theft? I would imagine it is in the single digits at the highest. This leads to most people reacting to car alarms with mere annoyance rather than any suspicion of an actual crime.
I don't know a real solution to this. It is really just human nature to react to personal experience and if that experience says alarms are usually wrong, we will begin to ignore them. Meanwhile the makers of the alarms have an incentive to be overly sensitive because a false positive is much less dangerous than a false negative. This results in a sort of tragedy of the commons in which all alarms end up being too prone to trigger accidentally.