Indeed, the internet may be our saviour here – but how people move about on the internet bears a lot of ressemblance to how hunter-gatherers look for food (can’t find the link), so it may not be that different. But you’re right, a new model would be good (though it still may be valid off internet). Anyone know of any new models exploring these phenomena?

The collective problem is harder to solve and we can probably only chip away at the edges. One possibility is to imagine cross-cutting communities, so that people are members of multiple communities and interest groups which don’t all divide up the same way. Both liberals and conservatives might enjoy sports, for example. Atheists and believers can both get involved with cooking groups. Promoting such “non-denominational” interest groups could be a way to bring these various communities together.

The more interests, the more diversity, the less clumping, good. But there may still be a risk – replace neighbours with “people I will talk about a subject X with”. If X is bias prone, we may still have the same problem; even if I’m meeting lots of other people, I need to talk about X with them. On top of “non-denominational” interest groups, we need to see that people do occasionally chat of the “denomination” issues at these groups – so, probably, the more sociable the interest groups, the better.

The collective problem is harder to solve and we can probably only chip away at the edges. One possibility is to imagine cross-cutting communities, so that people are members of multiple communities and interest groups which don’t all divide up the same way. Both liberals and conservatives might enjoy sports, for example. Atheists and believers can both get involved with cooking groups. Promoting such “non-denominational” interest groups could be a way to bring these various communities together.