From Wired Magazine:

The [peer review] process is lousy at policing research. Bad papers get published, and work that's merely competent (boring) or wildly speculative (maverick) often gets rejected, enforcing a plodding conservatism. It seems silly to say this about a system that's been in development since the mid-1700s, but the whole thing seems kind of antiquated. "Peer review was brilliant when distribution was a problem and you had to be selective about what you could publish," says Chris Surridge, managing editor of the online interdisciplinary journal PLoS ONE. But the Web has remapped the universe of scientific publishing – and as a result, peer review may finally get fixed.

In June, Nature began experimenting with a new method online. Authors submitting papers can choose a two-track process. While the work goes through the usual peer review drill, a preprint version gets posted on the Web. Anyone – even you – can comment, as long as you attach your name, affiliation, and email address. As of July, 25 articles had undergone this process, and the journal plans to issue a report late this year on how the test went.

… In other quarters, traditional peer review has already been abandoned. Physicists and mathematicians today mainly communicate via a Web site called arXiv. (The X is supposed to be the Greek letter chi; it's pronounced "archive." If you were a physicist, you'd find that hilarious.) Since 1991, arXiv has been allowing researchers to post prepublication papers for their colleagues to read. The online journal Biology Direct publishes any article for which the author can find three members of its editorial board to write reviews. (The journal also posts the reviews – author names attached.) And when PLoS ONE launches later this year, the papers on its site will have been evaluated only for technical merit – do the work right and acceptance is guaranteed. "Data becomes useful only if it's shared," Surridge says. "At the moment, our mechanisms for sharing information are the traditional journals, and if they're hard to get into, data is completely lost."