Besides helping the author to correct errors in a paper, peer review has traditionally had two benefits to the scientific community. One has been to control the costs of publishing and distribution of articles. The other has been to guide colleagues as to which articles are worth reading.

The internet has reduced the cost of publishing to almost nothing, making the first benefit of traditional peer review also almost nothing, except for journals still printed on paper. At the same time, the internet has increased the second traditional benefit of peer review -- helping colleagues decide what articles are worth reading. But that function could be served by a much more informal, distributed means than formal peer review. In fact the informal process would essentially revert to the old process of sending letters of recommendation to colleagues (just sent by email or posted on personal websites)!