From: Gary F. York (gfyork@mchsi.com)
Date: Fri Jan 05 2007 - 11:18:03 MST
Danila Medvedev wrote:
> As an economist with a PhD about discounting and project valuation, I
> find it highly disturbing that you are using such antiquated
> instruments as "interest rates" and "discounting".
I note that 'antiquated' is not equivalent to 'invalid' or
'inapplicable' or 'falsified'. But you do seem, later, to be suggesting
'inapplicable'.
I also note that the ideas of interest rates and discounting rely on a
few preconditions:
1. Entities that can have 'things' (even if only information)
2. Entities that have some way of deciding, at any instant, what is the
preferred use of the resources under their control (and some concept of
'private property' in which individual entities do have resources which
they control).
3. Entities that can engage in transactions or exchanges.
4. A suitable medium of exchange (someone suggested energy in this case)
5. Entities that prefer to have things sooner rather than later -- in
other words, they express 'time preference'.
It seems to me that the absence of any of these five conditions would
indeed invalidate the concepts of 'interest rates' and 'discounting'
and, unless I've failed to make my list exhaustive, the presence of all
five is sufficient to enable those concepts.
Discounting
'Discounting' comes about by our attempt to allow for time -- and 'time
preference'. It attempts to answer the question, "How much less is
something worth if I get it later rather than now?" It should be
obvious that if you have something that you are willing to exchange,
now, for $100 you will possibly be unwilling to make that exchange if
the $100 is to be received 10 years from now -- even if there is no
possibility of default. $100 to be received later is worth less, in
your estimation, than $100 in hand now. In other words, you 'discount'
the value of the time-delayed return.
(Clearly the concept transcends the specific medium of exchange.)
Interest
It can be argued that, at any instant, you attempt to choose the very
best use of the resources you control. (Of course, there may be little
agreement between individuals as to what constitutes 'best'.) If I want
to use some of your resources, I must induce you to part with them --
for a time. One of the most straightforward ways of persuading you is
to offer to return those resources to you -- later. With 'interest'.
How much interest I must offer will depend on:
1. How long I want the use of your resources
2. Your 'time preference' (how much more I must offer to return to you
simply to equal, in your estimation, the resources 'loaned' to me --
discounting)
3. Perceived risk -- of default and any other risks that you might see
eventuating from the loan
4. Expected return from your next best use of the resource.
>
> The whole discussion is completely wrong for the following reasons.
>
> It is not justified to speak about required rate of return for an
> entity which lives inside a computer. When most of the projects are
> carried out inside a virtual space (i.e. subjective time is not equal
> to astronomical time), when growth rates are high (and not limited by
> the same things as they are today), when the entities themselves live
> in a virtual space, when there might not be competition based on
> growth rates, etc. it is just stupid to discount future resource
> flows in the same way as we do today.
I don't see how anything you've said in this paragraph invalidates the
concepts of 'interest rates' or 'discounting'.
You dismissively introduce 'required rate of return' which is just a way
of applying the concepts of 'interest rates' or 'discounting' to
estimate whether a given effort will be productive over the time it
takes to exert that effort. And 'productive' is itself merely shorthand
for 'will yield more of value than is expended'. You suggest, by
implication, that an entity "which lives inside a computer" need have no
concerns about productivity. Why?
You assert that:
a) "When most of the projects are carried out inside a virtual space
(i.e. subjective time is not equal to astronomical time),"
b) "when growth rates are high (and not limited by the same things as
they are today)"
c) "when the entities themselves live in a virtual space,"
d) "when there might not be competition based on growth rates"
-then-
"it is just stupid to discount future resource flows in the same way as
we do today."
But none of your points, a through d, seem germane to the issue and I
cannot see that they even support your conclusion let alone require it.
>
> From an economist viewpoint, there likely will not be any discounting
> whatsoever, since what matters is total subjective time (i.e. amount
> of calculations), which is unrelated to astronomical time and is
> related to resources only. If a posthuman civilization is no longer
> worried about "real-time threats" (space aliens, asteroid danger,
> supernovas, etc.) that have to be countered quickly, it will probably
> concentrate on using efficiently all avialable resources for its
> purposes (no longer directly tied to the real physical world). I
> ignore all possibilities of final resolutions such as creating new
> universes for computation or dyson/tipler infinity scenarios.
It seems to me that the speculation as to how things will ("likely") be
very different in a virtual environment should probably not be made
unless you are first able to _prove_ that we are not _now_ living in a
'virtual' environment. :)
I readily grant that things _may_ be different, even radically so; but
unless one of the 'preconditions' I mention is absent it seems to me
that there must be some analog of 'discounting'.
Subjectively speaking, 'time' is just the observation that not
everything happens at once. The 'rate' at which things happen (and the
"amount of calculations") does nothing at all to alter the fact of 'time
preference': that entities generally prefer certain desirable things to
happen in their perceived _now_ rather than later.
Candidly, I am unable to discern exactly what point you're making in the
balance of your paragraph or even how it ties in to your prior comments.
Best,
G.
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:57 MDT