There seems to be at least as much variety of opinion now as before the web, probably more.

I don't think he's claiming the variation in content has decreased, but warns that dependence on collective rating systems could decrease the variance of an individual's effective sample size.

I also don't think the author claims the web has introduced any social problems, just that it isn't solving them yet. Are these even problems with technical solutions?

He seems to be ignoring the "long tail" argument that the internet, with its vast reaches, enables many people in weird niches to find one another.

I think I understand Jaron's premise - sites like facebook reduce the human element by forcing its users to add information according to a predefined structure. Presumably these same users may have created more free form web sites (and he equates this to individual creativity), but I don't really agree that it leads to group-think.

Facebook or any other site is just a medium - whether or not it leads to group-think is more a function of the participants than any conformity in their online 'meeting place'.

i bet he could roll a phat doobie.