Re: Variations on a theme:

Anders Sandberg (
Wed, 13 Aug 1997 17:11:50 +0200 (MET DST)

On Wed, 6 Aug 1997 wrote:

> Now my understanding of neurosurgery is limited at best...(I missed that day
> in Truck Driving School)...but from what I understand the nerve impluse is
> electro-chemical. That is it travels by ionic transfer thru the
> nerve(dendrite) and then by chemical transfer from axion to (?) ...across the
> gap. This obviously is speed limited. Several tens to several hundreds of
> feet per second. Pitiful.

Yes, but nature has done some optimizing. Along an unmyelinated axon
the signal is sent by ion channels opening and closing, creating an
electrochemical signal which moves fairly slowly. But if you
myelinate the axon, the signal can move as a purely electrical signal
(a change in trans-membrane potential) much faster, although it
decays exponentially. So most axons are covered with myelin except
at the nodes of Ravnier, where there are short stretches of
unmyelinated surface. The result is that signals "jump" from one node
to the other, and are amplified there so that they can reach the
next. Still limited, but faster.

> In an earlier reference it was mentioned that the ability had been aquired to
> fuse a chip to a nerve. A direct nueral interface on a single nerve level.
> Or perhaps I read it somewhere else....but I'm quiet sure that I DID read
> it.

Yes, there are such results. There was an article in Physics Review
(I think) about connecting a neuron to a field effect transistor.

> At anyrate. Imagine this: Using a micro-robot "surgeon" attach a fiber (as
> in "fiber optic") to a nerve near the top of the spinal column....then
> parallel that nerve along the spinal cord (thru the arms, legs...whatever)
> until it reaches it's destination...then tie in again. In effect cutting out
> several feet (possibly) of slow electro-chemical nerve impulse transference
> and replaceing it with light-speed. Does this suggest anything about
> increasing reflex time? Or more prosaicly,,,repairing spinal cord injury?

There is plenty of processing in the spinal cord (in the H-shaped
grey matter column along its length) which you would need to replace,
but your approach would work for much of the peripheral nerves. The
big problem is that our muscles are fairly slow too, so even if you
win a tenth of a second, the muscle speed would limit the use. But it
might be very useful for perception?

> But why stop there. Many of my favorite SF stories and novels use the
> device of a direct neural interface (David Weber-"Imperial Implants", Daniel
> Keys Moran-"Inskin", name but two) to directly control mechanisms or
> communications. Is this not possible now? At least in theory?

At least in theory it is possible. The big problem is doing something
useful with the connection, the brain is very "disorganized" (rather,
organized in a complex and individual way).

Anders Sandberg Towards Ascension!
GCS/M/S/O d++ -p+ c++++ !l u+ e++ m++ s+/+ n--- h+/* f+ g+ w++ t+ r+ !y