@colourincorrect Superrationality arises out of "know thyself" because an agent must first conceive of the self in order to conceive of the other. Only with knowledge of an other can the strategy of the other be taken into account.
The superrational agent follows the golden rule: "love the other as the self" in terms of "treat others as you would be treated".
I'm not sure this can be categorized as strictly altruistic because the motivation of a superrational agent is more correctly Pareto improvement, which does not require the condition of disinterest in, or diminishment of, the self.
@DukeZhou I find your analysis very interesting (esp. the association with Greek Maxims, though admittedly I know next to nothing about things of this sort)
" "love the other as the self" in terms of "treat others as you would be treated"."
^Reminds me of the notion of agape, which i recall was a form of mutually transformative (and self-improving) love
Again we see elements of super-rationality! How exciting!
I would be very careful about making explicit associations, but then again in some sense we are just cognitive agents playing an information incomplete game..... so who knows really?
Thanks for the good words! I'm confident philosophically in these two, primary Delphic Maxims because they are so fundamental and general. In some sense, minimax can be generalized to (-,+) and further to {-,+} which allows for maximin as well.
@colourincorrect Also, I'm an engineer as opposed to an academic, so I can go out on a limb with this stuff. Although I sometimes may do simple, mathematical proofs (such an analyzing an equilibrium in a game) ultimate validation is functional and algorithmic.
I stress philosophy because there are moral implications to game theory, and value alignment with strong algorithmic general intelligence, should it ever be possible, may be critical. Just doing the math or engineering without thinking about the social implications is highly problematic.
