Entropy and social groups

I sug­gest that there are de­fault pat­terns for so­cial groups, and they could be viewed as high en­tropy—what you’d ex­pect with­out know­ing more than that there was a so­cial group of a cer­tain size, pos­si­bly with some mod­ifi­ca­tions for tech level and sta­tus.

For ex­am­ple, I think that au­thor­i­tar­i­anism is the de­fault for gov­ern­ment—“we’re in charge be­cause we’re in charge, and it would be dan­ger­ous for any­one who tries to change that”. To­tal­i­tar­i­anism is lower en­tropy—it’s sur­pris­ing for the peo­ple in charge to have an ide­ol­ogy which re­quires them to make dras­tic changes.

The re­cent Elitist Jerks: A Well-kept Gar­den de­scribes an effort to fight one sort of en­tropy (the rep­e­ti­tion of the same ques­tions and an­swers) which re­sulted in an­other sort of en­tropy (an ex­ces­sively sta­ble and even­tu­ally frag­ile core group).

Main­tain­ing fun is an­other challenge in the keep­ing things al­ive cat­e­gory. Pleas­ant is rel­a­tively easy. Fun (which I’d say re­quires nov­elty) is harder, and I’m in­ter­ested in com­ments on what it takes to keep the fun go­ing.

There’s a the­ory that life ex­ists as chaos on the bor­der be­tween or­der and ran­dom­ness—I find this plau­si­ble, and it’s a differ­ent an­gle for look­ing at the Friendli­ness prob­lem. How can a sys­tem be built which con­tinues to per­mit (or even en­courage) in­ter­est­ing sorts of change, with­out per­mit­ting change so dras­tic that we as we are now wouldn’t rec­og­nize the out­come as still re­lated to us?