Could a computer write a paper that gets accepted?
Rooter: A Methodology for the Typical Unification of Access Points and Redundancy
MIT students wrote a program that generates random computer-science papers. One got accepted to a conference.
Three MIT grad students built SCIgen, a tool that spits out random but convincing-sounding papers, complete with graphs and citations.
They submitted one to a conference that was known for spamming researchers, and it was accepted. They raised money to go and give a (also generated) talk. SCIgen papers have been caught in real publications ever since.
Peer review, meet randomness. Randomness won.