For me, PubMed Commons came out of nowhere.  I was aware of some innovations in publishing (check out journals like Cryosphere for examples of progress in academic publishing) but to have a huge group like PubMed involved in pro-actively pushing boundaries is a real game-changer. Here’s why it’s so important: academic publishing as it currently exists is broken. We have been using a model for publishing that is built around a century-old method for the dissemination of information.  This involves (i) the submission of articles to an editor, (ii) the selection of a small number (usually 2-3) of referees to review the paper and make sure it is adequate, and (iii) a judgement made by the editor and the referees as to whether or not the paper should be accepted.  At that point, the paper is either published (in which case it becomes a matter of record) or rejected (in which case it is never heard of unless published elsewhere).  What this means is that ENORMOUS amounts of scientific information is never seen, and that information that is released in given a sometimes-cursory review by a small number of people who may not be experts in the area.  The internet should already have changed that in a number of ways:

1. Pre-review preprint servers – The mathematics and physics communities have been using pre-print servers for many years, almost to the exclusion of regular journals.  Arxiv is the gold-standard, which holds papers that have not been reviewed in a public forum where anybody can read and comment.  This removes the issues with limited peer-review so that all work can be evaluated by the wider community, made possible by the access to information provided by the internet. I was delighted to hear (just a couple of days before the launch of PubMed Commons) that the biological sciences are going to have their own pre-prints archive at Biorxiv. No word yet on when that will launch, though.
2. Post-publication comments – Once a paper is published, that paper tends to remain as a part of the scientific record.  Very few papers are ever retracted, even if deemed incorrect or fraudulent, and much of the critique that happens today takes place on blogs.  Journals have means for comment on articles, but these rely on standard publishing models (submit to editor, review, decision over publication).  However, PubMed Commons allows anyone (right now only PubMed authors during the pilot) to comment on any paper in the PubMed archive (which is over 23 million citations).  What’s more, the authors of the original papers can comment and update on work – particularly useful to link in subsequent papers or data that has been newly available where journals didn’t support the addition of supplementary data with the original paper.

All in all, exciting times and lots of progress being made!!