The lesson of part 1 was that the Net won't take care of itself. There is no nature that will protect cyberspace against change, and there is a tremendous amount of pressure for cyberspace to change. If the change continues along the lines it has taken so far, it will become a highly regulable space—not the locus of liberty, not a space of no control, but a technology of government and commercial power wired into every aspect of our lives.
But that's just one possible spin. The aim of part 2 was to suggest others. The built environments of cyberspace could be many, because the choices are many. And the potential power of government over these choices is great. Government has many tools with which to bend, or perfect, the architectures of this space. It has many techniques with which to make cyberspace what it wants cyberspace to be. These powers, however, are constrained—sometimes by law, more importantly by the code itself. Open code is one such constraint. But the possibilities of constraint are not themselves constraints. The message of part 2 was that little is determined, and much is possible. Choice is possible.
Part 3 exercised that choice. I began with a traditional legal technique for deciding how to go on—let the framers decide. Though the framers knew little about TCP/IP, the argument was that they established a tradition that can be translated into the context of cyberspace. They gave us the values, and our task is to carry those values into cyberspace.
But translation does not deal well with latent ambiguities. My argument in the balance of part 3 was that there are many such ambiguities. In four crucial areas of