[ExI] The Bomb verses a Email server

Anders anders at aleph.se
Wed Sep 14 23:38:45 UTC 2016

On 2016-09-14 18:16, Dan TheBookMan wrote:
> Has anyone here, aside from me, read Bertrand Lemmencier's essay 
> 'Nuclear Weapons: Proliferation or Monopoly?'? As one might guess, he 
> argues in favor of proliferation as lowering the likelihood of war. 
> One can imagine this working, but what about when proliferation 
> reaches the subnational level?


I took a glance, and was not super-impressed. There is a fun 
game-theoretic argument that things turn nice with many players, but 
this all assumes rational and error-free players. It also assumes 

It also seems to underestimate the number of possible conflict surfaces: 
among N players, there are N(N-1)/2 possible conflicts and if the 
probability of war between two players is p, then the probability of at 
least one war is P=1-(1-p)^(N(N-1)/2). If p is a function of N, to keep 
this probability below 1/2 p must decline as 
p(n)<1-exp(-2ln(2)/(N(N-1))) = O(N^-2).  Right now N=9, so if there is 
one new nuclear state the p better become 23% safer or the new state 
will reduce global safety.

The "more guns make society more polite" argument has never been very 
convincing with bombs.

Dr Anders Sandberg
Future of Humanity Institute
Oxford Martin School
Oxford University

More information about the extropy-chat mailing list