They've just found bacteria that make nanotubes, they make arsenic-sulfide nanotubes with unique properties.
A few years ago I did a bit of thinking about how to make bacteria
make carbon nanotubes. What you want to do is to find a secretory
system that exports some kind of molecule, like the bacterial
secretory system that some bacteria use to get rid of antibiotics.
Then, you change the inner pore into a machine that uses ATP to add
carbon atoms one by one onto a small nanotube primer. With this, you
could build nanotubes as long as you like, just keep feeding the
bacteria and let them go.
This is the future, and we're only a few decades away from this,
making biologically derived nanomachines. Real nanomachines (and
sub-nanomachines) aren't going to look like the machines we see
everyday, the atomic and quantum world is very different from the
macroscopic world, and we need to take our inspiration for
nanomachines from ideas that have been around for 3.5 billion years.