[ExI] neural interface reply

[William Flynn Wallace]

> I have to ask this question:  what do you expect a neural interface to do?

The literal answer to the question is simple:  Transduce neural signals to
and from electrical (or possibly optronic) ones, in a reliable fashion
without any adverse affects on the living tissue or the technological
components, and without disrupting the normal functions of the organic
parts.  We're still a long way from that, but progress is being made.

But I think your question was more along the lines of:  "What good would a
neural interface be?", and of course that's an open question, but here are
a few answers:

A fairly modest application would be to connect specific brain regions to
external circuitry for control of, for example, prosthetic limbs,
non-biological organs, new sensory equipment, etc.

Taking the idea further, and once the technology is worked out, you could
have very fine-grained interfaces that can talk to individual neurons and
feed the signals into/from detailed virtual-reality models of the body or
of parts of the body, so that your brain could 'drive' the virtual body.
 This would be good for full-immersion VR experiences, and also for
providing a control interface for new bodies.  Bodies which could be
anything from fully virtual to fully physical, any mix of the two, and any
mix of biological/technological for the physical side.

This will be difficult to do, and involve massive arrays of interfaces,
connecting to tens of millions of nerve-endings, but would be tremendously
useful.  You could, for example, take the central nervous system and
transplant it into a different body that was a hybrid of biological and
synthetic parts, and control it through a communications grid linked to the
interface.  You could then control virtually every aspect of such a body
(not to mention that the body could be designed and built to be vastly
better than a standard biological one).  And, yes, that's getting a bit
off-topic, but you can see the usefulness (or rather necessity) of neural
interfaces in a scenario like that.

All of the above presupposes that the biological brain will remain as it
is, producing your mind in the usual way, but there's more:

Taking things further - and later on, I imagine -, neural interfaces would
be useful in the process of 'gradual uploading', where rather than
replacing parts of the brain, or 'scanning' it, which are the usual
uploading ideas, we instead expand the mind into non-biological brain
machinery over a period of time, so that eventually the original biological
brain becomes only a small (and ultimately redundant) part of what produces
the mind.  At some point, you'd be able to ditch the original brain without
even noticing it.

Lots of assumptions there, but I don't think there's anything that's
theoretically infeasible.

Well that's a lot of ideas and no mistake, though I think I'd rather stay
all biological if I lived in the future.  All of this may be greatly
complicated by two recent findings:  glial cells, which outnumber neurons
in the brain, have now been implicated in lots more than their supportive
function:  they are involved in processing as well, and that opens all
sorts of ghost in the machine research.  Neurons may not do it all.  But
even if they did, many functions are not localized in the brain, so
connecting them with wires is going to be impossible.  I'd get rid of the
strong tendencies to cognitive errors and see if we need any more cognition
than that.  If AI fulfills its promise, we can let them do the processing
we can't, just like we do now with ordinary computers, only in a very
minimal way (i.e. they can't do anything we can't do except they can do it
faster).

And increasingly we are learning that the body helps to control the brain -
like the bacteria in our guts in very recent research.  bill
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20140314/b9488b55/attachment.html>