Peer review is a process that in the last 50 years it has become common in the scientific journals. However, although it is useful because it eliminates the manuscripts of low scientific quality, also tends to eliminate the very innovative ideas that could be important for the advancement of science. If we think that the quality of scientific writings have a Gaussian distribution then the peer review process eliminates both ends of the curve, very bad and very good, and only leaves the writings that are standard. In addition the ideas should be accepted or rejected by the entire scientific community, not only by two people.