Developing nanotechnology will be a major project -- just as developing nuclear weapons or lunar rockets were major projects. We must first focus our efforts on developing two things: the tools with which to build the first molecular machines, and the blueprints of what we are to build. This will require the cooperative efforts of researchers across a wide range of disciplines: scanning probe microscopy, supramolecular chemistry, protein engineering, self assembly, robotics, materials science, computational chemistry, self replicating systems, physics, computer science, and more. This work must focus on fundamentally new approaches and methods: incremental or evolutionary improvements will not be sufficient. Government funding is both appropriate and essential for several reasons: the benefits will be pervasive across companies and the economy; few if any companies will have the resources to pursue this alone; and development will take many years to a few decades (beyond the planning horizon of most private organizations).
We know it's possible. We know it's valuable. We should do it.