OSI network stack
The content-centric internet will require a new organization of the standard OSI model or network stack. The old Internet was a client-server model, the new internet is peer-to-peer. That implies radically different demands -- a more of a many-to-many network rather than a many-hubbed topology like the current internet. It's still many-to-many at the bottom (under layer 4), but now the session layers and above are also.
The important interface into this new Internet II is through the programming paradigm, not the networking one (which can stay, like I said in my message to IANA, in the lower 4 layers of the OSI model). In other words, the standard 7-layer OSI view of the Internet is wrong-headed and misinforms the intuition. Most content that is being generated now for the internet comes from users, that means that IP addresses are not matching content desires/providers properly because most of it is on user's machines. In fact, perhaps a better way to view Internet II is to turn the top 3 layers of the OSI model upside-down.
Rename the "Application" layer the Content layer and put it up against layer 4: transport. This will be the underlying content-centric layer. Then the "Session" layer is moved to the top, and becomes the "p2p" layer representing the individual user's current interest and interaction with the network for that session. The "Presentation" layer (#5), then, is the 3-d visualization layer, showing the user/session a view into the vast content repository, which stays more-or-less persistent due to the growing value relationships being made between all the knowledge and data on the internet (organized by the voting model into galaxies of order).