The notion of a content addressable network, where content is defined broadly enough to encompass services, is profoundly important. Some of us have pushed towards for years and it is nice to see some PARC weight behind it.

For an example of the profundity, consider that all location-addressed content is intrinsically ephemeral. When the ownership or maintenance practices of the location change, the content can change or disappear. With content addressability (in a uniform, stable, global way) we begin to have the possibility of authentic digital artifacts, addressable on the net by ordinary people in familiar ways (like following a link).

Or as another example of the profundity, consider the political, social, and economic power of ICANN - the ultimate sovereigns over the ownership of location names. Content addressability provides an alternative to submission to those or any other humpty dumpty when naming things for others to find on the 'net.

I dont have time to watch an hour long video. Can someone summarize? How do you address content without location? Is this the idea of p2p replicated content so that the network knows where all the copies of the content are? What happens if all the copies of the content are deleted from the network?