On Thu, 1 Jan 2004, Tim May wrote:
A few moments of thought will show the connection between replicators and general assemblers. A general assembler can make another general assembler, hence all general assemblers are replicators. And in fact this is necessary to make mechanosynthesis nanotech viable, as otherwise it takes all the multibillion dollar wafer fabs in the world, if they could make nanoscale things, to make some scum on the bottom of a test tube.
Or a few-dollar fermentation tanks with suitable bacteria, once its genome is tweaked in required way. Who ever said that the nanoparticles we need can't be proteins or organic molecules with required shape/properties? If viral particles can self-assemble from host-cell-synthetized proteins, if complicated structures like bacterial propulsion systems - or even whole plants - can be formed, why not nanomechanical systems? Why bother with assembling machines when they could be grown? I hope I didn't screw up my understanding of "nanosynthesis". If it is "build anything you want by telling the general assembler", then this won't work and would need a lab; but for mass-producing nnoparticles, eg. surface coatings or elements for camera or memory arrays, biotech should be good enough.