Somebody exposed Steven Den Beste to serial experiments lain. He's a cell network systems engineer, that shit is like catnip to network engineers. The end result is an exceptionally long essay for a famously long-winded essayist. The basic argument is that he's hepped up about the possibility that internet-enabled "hive minds" are going to achieve superhuman intelligence within our lifetimes - intelligences capable of thoughts impossible in individual human minds.
Surprisingly, he doesn't add the obvious caveat that what he's talking about isn't "group mind" - intelligence with personality. lain, after all, is a group personality. The characteristic of a hive mind is intelligence without consciousness. I suppose you could say that hive minds can be distinctive in a fashion similar to "personality", but it's clearly not the sort of personality that says "I" of itself.
Den Beste offers Open Source projects and blogspheres as examples of emergent hive minds. I don't have much personal experience with OSS projects, but I sort of see what he's talking about in terms of blogspheres. He makes a strong argument that a catastrophic-expansion, singular hivemind like Lain isn't the most likely development. It presumes a perfect, literally instantaneous matrix, I think. The current models of hivemind are very much task-oriented - specific long-term projects, ideologies, affinities, etc. I have to wonder if general-purpose internet hiveminds are even theoretically possible, let alone likely. This is why I've been somewhat hostile to the notion of The Blogosphere, as opposed to an array of lesser blogspheres. The blogspheres are discontinuous. Hrm, I wonder if anybody's done degrees-of-separation studies to see what the average associative order is? I mean, six degrees is obviously too low, but it's certainly not a single-degree situation.