Nooron is free (LGPLed) software which aims to provide a platform for collaboration on a global scale. The theory behind this effort is that a way to achieve such inclusiveness while also promoting quality is to create the mother-of-all-peer-review-systems and apply it to flexible, distributed knowledge representation. Unlike the majority of collaborative filtering systems, Nooron is based on user-extensible criteria (dimensions along which evaluations – ratings – may occur.) An unprecedented (except in biology) level of self-organization is expected to result from diligent application of criteria-based filtering to nearly every aspect of the system. Here's a simple inductive argument for suspecting that a cascade of amelioration and diversification is likely. Start off with a distributed knowledge representation system into which ontologies as well as application components are thrown. Apply a general purpose, omnidirectional improvement mechanism to all the contents of the system. Where does it stop?