Re: hating the creator

On Mar 20, 10:51 am, Mark Atwood <m...@xxxxxxxxxxxxxxxx> wrote:
"Greg Egan" <grege...@xxxxxxxxxxxxxxx> writes:

But unless
they're fundamentalist ideologues in love with conflict for its own
sake, they will notice that they can limit conflict over resources by
self-imposed constraints. This is not a small pay-off.

*Only* if they can either convience *everyone* around them to agree to
the same contraints and the same level of constraints, and have
everyone feel that the the constraints are locally and individually
fair (good luck with that), or if they can convience their neighbors
that they have an "over committment stratagy" such that trying to
encroach their boundary will cause costs and losses exceeding that can
be gained by the encroachment.

The second is more likely because it's decentralized and distributed
and doesn't require a universal in local space and time imposition.

In fact, it's probably the evolutionary discovery of overcommittment
stratagy that lead to the development of "terrority", "property" and
"mine" in the first place, with attendent advantages in lower costs
and losses in conflict.

These are political problems which might or might not be soluble.
Whatever is going on in the universe, at some level there will be
political problems which might or might not be soluble. Maybe we can
agree on that much and move on.

Perhaps the following will clarify my frustrations with this debate.
To a computer scientist, the most important aspect of any problem is
its scaling behaviour: given the n-sized version of the problem, what
is the function f(n) that describes the amount of resources needed to
solve the problem. (Normally we'd talk about two separate functions,
to describe the runtime requirements and the storage requirements, but
for my purposes there's no need to bother).

Now, if you tell a computer scientist that you are going to divide the
resources available to her to solve this problem by some constant c,
where c lies between, say, 2 and 10, she will not even blink. Why?
Because all she really cares about is the nature of f(n): whether it
is linear, polynomial, exponential, or whatever. She knows that
unless she's starting from the absolutely universally optimised
algorithm -- which for non-trivial real-world problems is vanishingly
unlikely -- the odds are overwhelming that a handful of very minor
innovations will allow her to make the transition:

f(n) -> f(n)/c

and she will gain an advantage that will completely neutralise her
initial loss. More importantly, if she makes even one *serious*
innovation -- say lowering the order of a polynomial, or the rate
constant of an exponential -- then not only will she neutralise her
initial loss, she will consider it utterly irrelevant. And if she's a
*really* good computer scientist, and the problem is a big, complex
juicy one, there's even the possibility that she will change its
complexity class: turning, say, an exponential f(n) into a polynomial
f(n). Dividing her resources by c at the start will then be a gnat's
wing against her bumper.

Now, suppose Greenpeace get wind of the coming age of thinking
machines, and call dibs on 1% of the universe as a wilderness
reserve. This might cause certain people to hyperventillate, but the
computer scientist who won the world-wide competition to be chief
architect of post-human civilisation will say: "Pah! Show me some
respect! Make it 99% of the universe!" Because she knows that the
qualitative nature of the problem, the scaling behaviour, is
completely untouched by this constant linear factor, and because she
knows that she and her mind children will be so smart that over the
next few billion years, they will innovate so many times that this 99%
will be one molecule on one torn scale of a gnat's wing.

.

Relevant Pages

  • Re: hating the creator
    ... they will notice that they can limit conflict over resources by ... the same contraints and the same level of constraints, ... Whatever is going on in the universe, at some level there will be ...
    (rec.arts.sf.written)
  • Re: Conflicting software / hardware
    ... A conflict that occurs when the same system resources have been allocated to ... Conflicts/Sharing is located in the Hardware Resources category in System ... If you have any problem devicewill be listed in Device Manager. ...
    (microsoft.public.windowsxp.help_and_support)
  • Re: SAS Soldier quits Army over Iraq -- Glowing refererence from regiment
    ... Denys a country of human resources that it would otherwise have. ... Conflict is the fairest and blindest judge of all. ... Conflict is the fairest judge of merit. ...
    (uk.politics.misc)
  • Re: Boots up in safe mode-graphics card message.
    ... resources would conflict with its resources. ... type that has a conflict and what it is conflicting with. ... listings eg in the device resources tab as described in the fix. ... Also my device manager does not show any Conflicting Device list as ...
    (microsoft.public.windowsxp.help_and_support)
  • Re: Set the project end-date
    ... constraints were probably set by typing in the task Start dates. ... Additional outsource resource for Development can be ... resources and no resources are over allocated. ... outsource to help shortend the project period. ...
    (microsoft.public.project)