John Clark writes:
> I found an article on the net about that breakthrough in molecular computing
"Breakthrough" is a rather strong word. We should reserve it for truly
special occasions. Demonstrating switching behaviour in an isolated
molecule alone is not new nor particularly useful. The hardly tackled
problem of building macroscopic blocks of designed circuitry by
autoassembly has been and still remains the bottleneck for real-world
circuits. Whether Langmuir-Blodgett 2d array memories or molecular
crystal 3d memories or even molecular circuit cellular automata
machines, the problem of hierarchical complementary-shape
self-assembly has not yet received as much attention as it
should. Without this a mole of the fanciest molecular switch will
remain a featureless coloured goo in a flask, not a machine.
> that I wrote about last week. The researchers sound astonishingly optimistic.
> Ten years still seems a little early to me, but I don't know, maybe Eliezer is right.
Oh, we should have molecular memories (probably 2d, as single-layer
2d-crystal assemblies of protein-enveloped molecular bits in
polymerized Langmuir-Blodgett monolayers on silicon substrate) 10-15
years from now. But these are not yet computers, only parts of
computers. Singularity will have to wait little a bit longer, I'm
afraid.
> John K Clark jonkc@att.net
>
> http://www.uclanews.ucla.edu/Docs/LSSW358.html
This archive was generated by hypermail 2b29 : Mon Oct 02 2000 - 17:36:31 MDT