By Jaron Lanier via BoingBoing
This is a definitely very thought-provocative piece, so are the comments in the Reality Club to the article.
I wrote about meta-meta-meta data over 1 year ago (I should finish the work!) noting that metadata is a "derivative" from the original data, usually with a purpose. "Metadata", as in Dublin Core, is for discovery of resources. So is inverted index used by major search engines.
Other metadata exist in many different forms to serve different purposes. RSS is another form serving the need of aggregators, which Jaron refers to collectives. To him, wikipedia is also a collectives, many anonymous authors aggregates information into a collection.
Metadata are themselves data and hence can further aggregate, recursive infinitum in fact.
Jaron warns about the lost of personality in the layers of layers of meta-ing. Using wikipedia as a specific example:
For instance, most of the technical or scientific information that is in the Wikipedia was already on the Web before the Wikipedia was started. You could always use Google or other search services to find information about items that are now wikified. In some cases I have noticed specific texts get cloned from original sites at universities or labs onto wiki pages. And when that happens, each text loses part of its value. Since search engines are now more likely to point you to the wikified versions, the Web has lost some of its flavor in casual use.
He also points out the success of some meta-blogs such as Boing Boing being
run by identified humans, which served to aggregate blogs. In all of these formulations, real people were still in charge. An individual or individuals were presenting a personality and taking responsibility.
In our field, we appreciate Stephen Downes effort because his OLDaily is a selection of pieces by him. We are looking at the world through a lens we have trust in.
In one way, aggregates/collectives are network, linking different pieces of information together. Collectives which carry "personality", such as Boing Boing and OLDaily have significant added values.
We get the maximum benefit from the collective only if we follow the links to the nodes and dig into the nodes. We learn from the nodes, NOT from the collectives. Collectives serve as a guide, a special viewing platform for us to see the world.
Learning is about experiencing the "stuff". Knowing-where and knowing-who are strategies that can help us find the "stuff" which may be of interest. We don't need to know everything to be able to do something, e.g. a pilot is not an airplane designer or a jet-engine engineer! We need the right level of metadata in order to satisfy our curiosity, need and work.