The New York Times ran an article on a new academic research project whose goal is to redesign the Internet from scratch. The most valuable product that will come from this effort is not new technology, but formally trained security researchers. This past weekend the old gray lady ran an article by John Markoff about how a group of universities received a large grant to draw up a brand new Internet from scratch. The motivation for the work is, as the group claims, that the internet in its current state cannot be secured, and a new internet needs to be drawn up to support security from the start.
The internet isn't insecure because the fundamental architecture is broken. The internet is insecure because people are using the internet using poorly secured systems and are willing to run any program they come across regardless of its purpose. This problem can't be solved on a protocol level without regulating what applications can be run and who is allowed to connect to each other, which risks stifling new technologies like peer to peer.
The biggest value that will come from this project is the production of any universities' primary output: new engineers and researchers. By allowing students to spend a few years thinking about what could be possible in security when they start with no constraints, they may come up with an innovation that can be bolted onto a current technology. When they graduate, they can either bring their research to market or train new engineers and researchers, allowing everyone to gain value from their grant.