The more I thought about it … it seemed like rational agents couldn’t trust anyone (the best course is to convince them to trust you and then betray them while never trusting anyone yourself) except in the early and middle stages of iterated games. But a society where everyone irrationally trusted everyone else, and irrationally nobody betrayed anyone else, would be more successful than the ‘rational agent’ community. (all things being equal; if their irrational trust also caused them to irrationally trust lions...) It might stretch the word evolution too much, but I think the term “competitive selection” applies to this process of societies competing with each other for growth and the most effective societies wiping out the less effective societies (wiping out or completely integrating, as the lesser society’s land and resources would already be purposed towards supporting a society, and therefore more desirable than land requiring work).
Basically, what if ‘trust’ is because a society where everyone trusts the other guy to cooperate in a PD was successful enough to dominate the landscape?
NB: Originally I had thought of trust as a sort of greenbearding. Is there an analogous concept in sociocultural evolution?
A population or society in which everyone trusts completely is not an ESS. A population or society in which everyone adopts the slogan “trust, but verify” and cooperates in the punishment of defectors and non-punishing freeriders probably is an ESS, assuming the cost of verification and punishment are low and verification is reasonably effective.
it seemed like rational agents couldn’t trust anyone (the best course is to convince them to trust you and then betray them while never trusting anyone yourself) except in the early and middle stages of iterated games.
In the real world, the iteration never completely ends.
Okay, not evolved adaptations, but how about culturally/socially imprinted cognitive biases? Something about
clicks with Nancy’s comment here.
The more I thought about it … it seemed like rational agents couldn’t trust anyone (the best course is to convince them to trust you and then betray them while never trusting anyone yourself) except in the early and middle stages of iterated games. But a society where everyone irrationally trusted everyone else, and irrationally nobody betrayed anyone else, would be more successful than the ‘rational agent’ community. (all things being equal; if their irrational trust also caused them to irrationally trust lions...) It might stretch the word evolution too much, but I think the term “competitive selection” applies to this process of societies competing with each other for growth and the most effective societies wiping out the less effective societies (wiping out or completely integrating, as the lesser society’s land and resources would already be purposed towards supporting a society, and therefore more desirable than land requiring work).
Basically, what if ‘trust’ is because a society where everyone trusts the other guy to cooperate in a PD was successful enough to dominate the landscape?
NB: Originally I had thought of trust as a sort of greenbearding. Is there an analogous concept in sociocultural evolution?
A population or society in which everyone trusts completely is not an ESS. A population or society in which everyone adopts the slogan “trust, but verify” and cooperates in the punishment of defectors and non-punishing freeriders probably is an ESS, assuming the cost of verification and punishment are low and verification is reasonably effective.
In the real world, the iteration never completely ends.