My favorite part, and the reason I'm talking about it here, is the idea that automated methods of posting weblog comment spam will get even more crafty, and taken to its logical extreme, a 'bot' will actually pass the Turing test in this limited context -- we'll be having meaningful discussions with these bot programs!
Someone else points out that, as long as they're well-behaved and actually contributing to the conversation, who cares?
Despite what Jeremy says, in his very judicious and proper tone, I do insist and concur with Avram and others: this is evolution in the best sense:
spambots create garbled test, which we delete.
More spambots come, v1.2, that create slightly less garbled test and on-topic posts. We delete.
Spambots v3.4 attack with their witty and accurate vision of the blogosphere, having traversed it all, and are able to debate our points and make a coherent argument.
We are deleted.
The main blog entry closes with a Heinlein reference: Indeed, are we sure it isn’t already happening right now? I mean, I know I’m real, but where did all you zombies come from—?