This is going to be a pretty technical post, but I do think the idea of what's going on is fairly cool. In layperson's terms, there are small pieces of DNA called LINEs in the human genome which carry the information necessary to move a copy of themselves to a new location in the genome. You could think of it as a short stretch of computer code which includes instructions to copy itself elsewhere inside a software program. This jumping, or "retro-transposition," by LINEs occurs very rarely in normal cells, and is basically like a parasite inside your DNA.
The really cool part about parasitic interactions is that they frequently are under evolutionary pressure to become symbiotic relationships. In situations of mutual benefit, the host then has less of a burden, and the parasite gets a happier host. A recent paper in Trends in Genetics shows that some LINEs have actually been subverted by their mammalian hosts into an instant regulatory mechanism. LINE derived micro-RNAs (miRNAs) can dock to existing LINE integration sites, which are frequently found in the 3'UTR of RNAs. This is like adding a new control button (if the host can harvest the opportunity).
miRNAs, which seem to be the topic du jour in the genomics world, are very short pieces of RNA which can control several aspects of regular messenger RNA biology including translation into protein and subcellular localization (for clarity, I will refer to messenger RNA as "messages"). miRNAs accomplish this by docking to very short sequences, which typically fall in the 3' untranslated end (the RNA sequence after the part which codes for protein) of the message. A given miRNA can recognize many target messages, thus bringing a whole set of genes under a single control.
As it happens, the 3'end of messages is also a region where LINEs can insert without damaging the protein, and this in fact is very frequently observed.
What is new about this paper is that mammals have evidently turned parasitic LINE integrations into an opportunity for transcriptional control, by converting sister LINE sequences to make a LINE-derived miRNA. Because the parasite copies are distributed in many places, the LINE-derived miRNA instantly matches a whole set of targets. And because the miRNA machinery is well established for other short pieces of RNA, the whole control module can be plugged in to an existing infrastructure. What was a bug, has become a feature.