[extropy-chat] singularity conference at stanford

[spike]

> bounces at lists.extropy.org] On Behalf Of Jef Allbright
> Subject: Re: [extropy-chat] singularity conference at stanford
> 
> On 5/15/06, "Hal Finney" <hal at finney.org> wrote:
> > I remember back when Hofstadter was doing his column for Scientific
> > American, in the 1980s...

Yes I was in college then, and part of the crowd that hung on pretty much
any word that Hofstatder wrote or uttered.  I remember that so well because
I was coming up on graduation when the June 83 SciAm came out with the
following entertaining game.  Scientific American was offering a million
dollar prize in a lottery wherein the entries were free, just send in a
postcard with your name.  (This was before email was common.  We computer
geeks had it, but most did not.)  You could enter as many times as you
wished, in fact there was no need to send a bunch of cards, a waste of paper
and postage, just write the number of entries on your card.  Of course your
chances of winning would be proportional to your number of entries, as in
any lottery.  The catch was that the million dollar prize was divided by the
total number of all entries.

Hofstadter reasoned that his super-rational readers would perform some kind
of super-rational calculus, with the result that in most cases the reader
would not enter at all.  A handful of defectors would bring down the prize
to an affordable level.  What he did not count on is that among the
super-rational are the super-defectors.  A super-defector not only wants win
at all costs, he has as a second goal to keep others from winning.  He
received cards with a 1 followed by the rest of the card filled with zeros.
He received cards with things like 9^9^9, far more than the number of atoms
in the universe, which of course made the prize zero, even if it were
somehow possible to determine a winner.

This must have been discouraging to him, for the million dollar lottery game
demonstrated the long term futility of the policy of mutual assured
destruction as a means of maintaining nuclear peace over the long run.  This
experiment suggests that as more governments get nuclear weapons, eventually
we get a nuclear-armed leader whose actual goal is not just to protect his
own country, but to destroy other countries.  I think of this nearly every
day, as I read of the comments made by a certain middle-eastern president.

spike