So yesterday, I was dreaming about nanotech.

Specifically, I was dreaming about Shelly, datan0de, and nanotech. In a weird kind of way.

In the dream, the three of us were in a huge industrial factory of the kind that was built at the start of the industrial revolution–an enormous brick building, with spinning metal shafts along the ceiling that drive great pulleys connected by long belts to machines that did the actual work (and sometimes lopped off people’s fingers). For some reason I’m not quite clear on, it was desperately, vitally important that we come up with a way to construct generic, programmable nanoscopic assemblers, soon (like in six months), or we would all die.

During the dream, the three of us were arguing about the best way to go about constructing prototype assemblers. datan0de, who was wearing his DragonCon gear, was arguing for a “bottom-up” approach: design a molecule that would do what we wanted, then figure out a way to make that molecule spontaneously self-assemble when the correct ingredients were brought together in the correct way.

I was arguing for a “top down” approach, where we first design the assembler, then construct it on a molecule-by-molecule basis by using devices such as atomic force microscopes to place every molecule in position. Once the first assembler had been hand-built in this way, we could then program it to build more assemblers.

Shelly was arguing for a third approach–namely, designing the assembler, then constructing protein-based machinery to build the assembler by using custom strands of DNA to create the protein-based machinery. She argued that datan0de and I were trying to reinvent the wheel; the mechanism to create proteins from a strand of DNA already exist, so if we could figure out how to build a machine out of proteins that would construct the assembler, then we could use the machinery inside a living cell to cause the cell to churn out assemblers just by cooking up the appropriate DNA and introducing it into the cell.