gwern comments on Re-formalizing PD - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (57)
As a functional programmer, I can't help but notice that invisible global state like
randomdestroys the referential transparency of these programs and that they cease to be functions. If the random number generator's seed were passed in as an argument, restoring purity, would termination be restored?(Offhand, I don't think it would. Even if program A used each digit to decide whether to defect or cooperate, program B could just follow the same strategy but choose the reverse and simulate one more step.)