Academia in trouble

Came across the SCIgen page at MIT, complete with its own fantastically hilarious blog. This is a computer program that generates computer science papers randomly. The three students behind this program even got a paper accepted at a conference, and then went to the conference where they gave three randomly generated talks.
I am not sure what this indicates, other than that the “publish or perish” system seems to generate some seriously deficient reviewing. There seems to be a market out there for weird conferences (always held in attractive locations) with vague (but long) titles, as well as truly obscure journals. Time to clamp down at some point?
Anyway, kudos to the SCIgen guys – now all they need to do is get published, as was done in the Sokal affair. Wonderful.