>>13973095Many of you don't get the premise at all, it is not "ooga booga strong Greg bonk weak Greg" or "bong first win bumfight". It is more epistemically prudent than that. The state for selection it provides comes from a game theoretical approach for the agents in question that you can demonstrate.
(1) We have the capability to first strike another agents to debilitate their first strike capability if present, of not also the capability to annihilate
(2) We can presume they may also have any capability we have
(C) We can presume they may have the capability to strike us
We don't know, but we can relatively safely guess at some point 1 will hold. We already can destroy ourselves, it is not hard to imagine the ways we can extend this. 2 may hold for all sort of reasons even if you don't assume technological development being relatively uniform.
So given this we have this game with strategies
S = {CS, CS', C'S', C'S}
For all players where C is contact and S is strike and their primes are don't contact and don't strike.
The argument is that CS' needs CS' from the other player to win, which even with all options being equally likely is very low probability. We simply do not know which options are likely, but under the dominance principle you assume others assume XS.
In fact, arguably C'S when possible, since that practically ensures a first strike.
