> Hence I think that people's reluctance to "give
> themselves $1 in the past" is more due to the difficulty of analyzing
> such an impossibly hypothetical scenario.
I agree that the weakest point of their argument is that it is
difficult to tell how much people value their past selves.
> > It is fundamentally not an issue of learning. It is a conflict of preferences.
> I think that constantly regretting most past actions, without changing
> our current behavior, can fairly be called a failure of learning.
Imagine we were in a club with a rotating chairmanship, and that the chairman
decided where we would meet each month. Then each month we might meet
at that month's chairman's favorite restaraunt. If we all got together on voted
on where to meet, however, we might compromise on some place and then get
the benefits of having a regular meeting place. And on average we might
each prefer the regular meeting place compromise to each having our favorite
choice every few years, and then having to go to a different place each month
at other times.
Similarly, regretting our past actions can be a matter of failing to coordinate
among our different selves at different times.
This archive was generated by hypermail 2b30 : Mon May 28 2001 - 09:50:19 MDT