[ExI] Affecting Past Experience
lcorbin at rawbw.com
Wed Jul 25 06:05:11 UTC 2007
>> I don't really know why you care more about these discrete *future*
>> days than you do about the *past* ones. If tomorrow is a day that
>> gets bombed..., then you'll simply not
>> experience that. But the day after tomorrow and all the days
>> beyond, it will be exactly the same as if tomorrow happened.
>> (Actually, were I scheduled for root canal surgery for tomorrow,
>> then I would even hope that it got bombed!)
> This brings up another twist in that you wouldn't know if every second
> day of your life, or an even larger proportion, did not actually
Yes. If we grant that "knowing" is an act requiring runtime, then
to know that a past day really occurred (or *assume* that it did
where assuming is also a thinking type experience) does, of course,
necessitate that I get some "on" days (to use your term) to reflect
> Would you agree to be given some advantage on your "on" days if
> you gave up a proportion of your future runtime as "off" days?
Yes. I'd try to frame the calculation (estimate/guess, really) in terms
of total benefit I receive. Some days are neutral, in that it seems I would
have been no better off and no worse off if they had not actually been
experienced by me (i.e., did not actually) occur, with me retaining only
the (false) memories of them.
Other days are so positive that they deliver as much benefit as many,
many mediocre days.
>> > In fact, I would prefer that all my past days be destroyed if it could
>> > save one future day, even though that way the total runtime of all
>> > instances of me is reduced.
>> Evidently you and I simply calculate these things differently. I would
>> not like losing some actual very pleasant day in my past (retaining
>> only the memories). True, I cannot tell if it gets bombed or not,
>> but I have the intellectual knowledge that my life as a whole was
>> worse off for not having really experienced that day.
> The intellectual knowledge would be worth something, but it wouldn't
> be worth as much as actual future experiences. Would you give up the
> last ten years for the next ten years, if your and everyone else's
> memories of the last ten years remained unchanged?
Again, I guess I just don't see actual future days being any more valuable
than actual past days. The tendency, yes, is to favor future days because
we get to anticipate them, but I myself reject anticipation as a solid
foundation for making decisions.
Mike Dougherty had written
> I was thinking about your Anticipation Dilemma. If you should be eagerly
> anticipating the dinner you ate last night because some earlier point-in-time
> version of yourself has not yet eaten it and you must embrace the
> anticipation of all your selves.
and I should hasten to add that this characterization, although informally
valuable, does lack the feature of my more complicated thought experiment
because past selves are not normally memory supersets. In my original post
which lays out the whole thing, quite a bit of juggling and not a few assumptions
were necessary to get a memory superset into the actual past.
More information about the extropy-chat