It is hard to imagine a more stupid or more dangerous way of making decisions than by putting those decisions in the hands of people who pay no price for being wrong.
It is hard to imagine a more stupid or more dangerous way of making decisions than by putting those decisions in the hands of people who gain nothing from being right.
Or:
It is hard to imagine a more stupid or more dangerous way of making decisions than by putting those decisions in the hands of people who have no strong reason to prefer the world in which their decisions are right, over the world in which they are wrong.
I think the “pay no price for being wrong” formulation is stronger than the “gain nothing from being right” one because of loss aversion (which makes penalties a stronger incentive), and either is stronger than your second suggestion because of pithiness.
My take on it: I’d noticed that “people who pay no price for being wrong” primed ideas of punishment in my mind, not just loss. “People who gain nothing from being right” primed ideas of commerce or professionalism — an engineer gains by being right, as does a military commander, a bettor, a venture capitalist, or the better sort of journalist.
And the third formulation doesn’t prime anything but “this sounds like Less Wrong”.
The biggest problem with your first alternative is that in it, not having an opinion is equivalent to being wrong.
A lot of the problems with the financial collapse was that various entities and people got to play with the money of other people, with good payouts if they get it right, but no commensurate hit if they got it wrong. While the best outcome is still being right, this kind of situation is bad because it incentivizes taking risk over not taking it. So, a lot of people making those decisions loaded up on as much risk as they could take, ignoring the downsides.
If a doctor uses the established thresholds and refers a patient to surgery that turns out to be unnecessary, there are laws preventing that patient from suing her. The same is true if all the tests came back below the threshold, she said he was fine, and he later turned out to be super unlucky and have a totally undetectable form of cancer. The doctor did everything right. She just got unlucky. Those laws are really good. If they didn’t exist, it would either be impossible to practice medicine, or else doctors would be optimizing for not being sued rather than for doing good medicine even more than they already are. If they only existed in one direction (eg doctors who did unnecessary surgeries couldn’t be sued, but doctors who missed cancer could), that would be even worse—any doctor not heroic enough to go against her own self-interest would refer every patient to surgery.
-- Thomas Sowell
Interesting to contrast the connotation with:
Or:
I think the “pay no price for being wrong” formulation is stronger than the “gain nothing from being right” one because of loss aversion (which makes penalties a stronger incentive), and either is stronger than your second suggestion because of pithiness.
Good points.
My take on it: I’d noticed that “people who pay no price for being wrong” primed ideas of punishment in my mind, not just loss. “People who gain nothing from being right” primed ideas of commerce or professionalism — an engineer gains by being right, as does a military commander, a bettor, a venture capitalist, or the better sort of journalist.
And the third formulation doesn’t prime anything but “this sounds like Less Wrong”.
The biggest problem with your first alternative is that in it, not having an opinion is equivalent to being wrong.
A lot of the problems with the financial collapse was that various entities and people got to play with the money of other people, with good payouts if they get it right, but no commensurate hit if they got it wrong. While the best outcome is still being right, this kind of situation is bad because it incentivizes taking risk over not taking it. So, a lot of people making those decisions loaded up on as much risk as they could take, ignoring the downsides.
I can imagine one easily. Where they have an active incentive to be wrong.
I dunno—Yvain here seems to have a good point:
[emphasis as in the original]