Re: Responding to Pre-dawn Unannounced Ninja Raids

Perry E. Metzger wrote:

> > There are surely alternatives to the extremes of unilateral
> > disarmament and an arms race? For instance, slowing down  the
> > race?
> 
> Don't ask the question as though it is theoretical. Don't try
> answering it as though your personal values have any bearing, because
> there is indeed an objective answer here. Try doing some game theory
> simulations and see how well unilateral disarmament works.

Perry, just out of curiosity which type of game theory system are
you refering to, the standard, single shot game thoery or the 
communicative model of Alker, Hurwitz and Rothkin?

As a computer scientist I would have thought you would have been 
up on the idea of adding memory to game theory interactions. In such
cases the optimal outcome can turn out very differently than in the
standard model. 

Claiming certainty from theoretical results is in general not a
good idea. Unless you can explain the relevance of the theory to a 
situation and explain why the assumptions in the model are justified 
you are not saying very much.

The fact that the US and the USSR did manage to negotiate disarmament
despite the standard game theory predictions shows that the system
is somewhat more complex than Perry's ideological view.

Briefly stated in Alker-Hurwitz a "memory" component was added into
the model. The prisoner's dilema was repeated on many occasisons in
a variety of contexts, including computer simulation and in practice.
In practice the prisoner's chose the joint optimal solution the 
majority of the time. 


>Oh, and don't give us stuff about how humans are above evolutionary
>pressures or nonsense like that, because we aren't any more above such
>pressures than we are above the laws of physics.

The "laws" of social scienst are not the "laws of physics". The "laws
of physics" aren't so constant either. Theoretical results should 
inform the intellect not serve as a substitute for it.

If you apply genetic programming techniques to the system the strategy
that evolves is typically a cooperative one. The facts is that the
theory applied in an evolutionary context disproves Perry.


		Phill

Hallam-Baker writes:
> As a computer scientist I would have thought you would have been 
> up on the idea of adding memory to game theory interactions. In such
> cases the optimal outcome can turn out very differently than in the
> standard model. 

Memory is the only way that things like iterated prisoner's dilemmas
become interesting.

> The fact that the US and the USSR did manage to negotiate disarmament
> despite the standard game theory predictions shows that the system
> is somewhat more complex than Perry's ideological view.

I'm afraid, Phill, that you didn't read what I said. Unilateral
disarmament is stupid. Multilateral is not necessarily stupid.

> >Oh, and don't give us stuff about how humans are above evolutionary
> >pressures or nonsense like that, because we aren't any more above such
> >pressures than we are above the laws of physics.
> 
> The "laws" of social scienst are not the "laws of physics".

Try creating a breed of Humans that don't want to have children and
see how many generations you can get them to live for.

Try creating a breed of Humans that like walking in front of cars and
see how long they last.

There is a reason humans do things like agressively defending their
children with their lives if need be. There is a reason humans resort
to violence when their place in the gene pool is threatened.

Some of this stuff is plumb obvious to anyone with half a brain,
Phill.

Perry

>Memory is the only way that things like iterated prisoner's dilemmas
>become interesting.

Having spent the weekend with Alker I'm hardly going to argue.

>There is a reason humans do things like agressively defending their
>children with their lives if need be. There is a reason humans resort
>to violence when their place in the gene pool is threatened.

That is one effect, but not the only effect. Depending upon what 
the conditions you set up arround the problem you can change
the outcome. What are the risks of fighting for example? Evolution
does not uniformly favour hawks, in terms of numbers the doves 
win.


>Some of this stuff is plumb obvious to anyone with half a brain,
>Phill.

Ah yes, and since iterated prisoner's dilema games are as
computationaly complex as the Mandelbrot set (the generator of the
Mandelbrot is in fact simpler), presumably you can calculate the
Mandelbrot set in half your brain Perry?

If it was "plumb obvious" it wouldn't take MIT profs to work
it out Perry. The world is far more complex than your simplistic
notions make out. Just because you can identify ONE effect does
not mean that you have identified ALL effects or even that you
have identified the dominant one. 


Your analyses are almost always junk because you only analyse one
side of the argument and deny that there is another side. You are
great at preaching to the choir Perry, problem is that you don't
convert anyone who isn't already converted.


		Phill

hallam@Etna.ai.mit.edu wrote:
> Ah yes, and since iterated prisoner's dilema games are as
> computationaly complex as the Mandelbrot set (the generator of the
> Mandelbrot is in fact simpler), presumably you can calculate the
> Mandelbrot set in half your brain Perry?

AFAIK, no one can "calculate" the Mandelbrot set precisely. There are 
many points about which you can't say for sure whether they belong to 
the set or not.

	- Igor.