One obvious problem with your predicted “good king” scenario is that a high rank in the “pecking order” inherently attracts badactors to it—which in turn are precisely the agents who will use that rank to do the most damage, both to other actors within the group and indeed to the organizational goal itself! Separating “pecking order” and “decision-making order” would seem to be the right answer—except for another wrinkle, that is; it seems that, among the few ways we know about of semi-reliably screening off bad actors, (1) requiring proof of having reached good decisions recently, especially on non-trivial and long-term matters; and (2) giving a boost to “prophet” types who can provide strong, complex signals of pro-sociality, not just momentarily but on a somewhat long-term basis (And yes, these reliable signals do seem to exist. For example, sound goal analysis that does not reduce to mere politicking and is based on a principled assessment of the group—this very blogpost provides us with a fine example! - impressive art, high-quality work in general—even something as simple as humour, perhaps!), are especially important. And both of these concerns would seem to push into the other direction, of commingling the “pecking order” and “decision-making” hierarchy to some extent!
I think it would be interesting to try and design honeypot hierarchies, that are expressly intended for bad actors to have harmless fun in, without dealing extensive damage to others. But a pecking order is not that; being low in the pecking order, especially with someone malicious at the top, is really bad. Thus, arguably, this is a goal that’s best pursued by the market system as a whole, not by a small-scale social structure that—like all social structures—comes with inherently “soft” and “pliable” incentives that will never manage to keep the most toxic agents on the shortest leash.
One obvious problem with your predicted “good king” scenario is that a high rank in the “pecking order” inherently attracts bad actors to it—which in turn are precisely the agents who will use that rank to do the most damage, both to other actors within the group and indeed to the organizational goal itself! Separating “pecking order” and “decision-making order” would seem to be the right answer—except for another wrinkle, that is; it seems that, among the few ways we know about of semi-reliably screening off bad actors, (1) requiring proof of having reached good decisions recently, especially on non-trivial and long-term matters; and (2) giving a boost to “prophet” types who can provide strong, complex signals of pro-sociality, not just momentarily but on a somewhat long-term basis (And yes, these reliable signals do seem to exist. For example, sound goal analysis that does not reduce to mere politicking and is based on a principled assessment of the group—this very blogpost provides us with a fine example! - impressive art, high-quality work in general—even something as simple as humour, perhaps!), are especially important. And both of these concerns would seem to push into the other direction, of commingling the “pecking order” and “decision-making” hierarchy to some extent!
I think it would be interesting to try and design honeypot hierarchies, that are expressly intended for bad actors to have harmless fun in, without dealing extensive damage to others. But a pecking order is not that; being low in the pecking order, especially with someone malicious at the top, is really bad. Thus, arguably, this is a goal that’s best pursued by the market system as a whole, not by a small-scale social structure that—like all social structures—comes with inherently “soft” and “pliable” incentives that will never manage to keep the most toxic agents on the shortest leash.