Re: a new thread

From: Ian Field (
Date: Sat May 13 2000 - 01:16:47 MDT

----- Original Message -----
From: "Anders Sandberg" <>

| > Any thoughts on human genetic manipulation? Say, with the purpose of =
| > creating a workable interface between an electronic microprocessor and a
| > human mind? Throw in some nanotechnology for physical interaction... =

| How would genetic modifications of the genome create a good
| brain-computer interface?

Earlier, I equated the concept to a structure similiar to the optic nerve.
I now believe a better analogy would be the speech centres of the brain. A
new, specialized, neural structure along similiar lines would probably be
appropriate - one suited to the purpose of interacting with one or more
complex, high speed microprocessors. Think of it as a client, or interface,
into the microprocessor's "human" API. The point is not to augment the
brain to be able to handle computer instructions, but rather to augment it
for more efficiect bidirectional communication with the aforementioned
computer(s). Most of the complex logic for interaction would be handled by
the machine (as in real software) - the human part of the equation is simply
communications. These are obviously rough ideas, and I don't have the solid
background in genetics/neurobiology that Anders obviously does...

| I think we transhumanists have to be careful about our enthusiasm and
| think carefully about the relative timeframes and levels of comelexity
| needed to achieve certain things. Setting upp chemical gradients in
| the brain so that engineered neurons growing on animplant can reach
| the right areas seems doable in the relatively near future; creating a
| nice socket on the skull through genetic engineering is much, much
| harder (requires us to master localized morphogenesis). And of course,
| nanotechnology is something completely different.

Here, you're jumping to conclusions based on my very brief initial
statement. I agree that a genetically hardcoded "socket on the skull" would
be impractical, ineffient, aesthetically displeasing, inflexible and
generally abhorrent. Consider instead the speech centres; responsible for
translating patterns of light and sound into recognizable values, and vice
versa. Is it not feasible that an enhanced speech centre could translate
electronic signals along those same lines? Is it also impossible to imagine
general nervous system enhancements which could help deliver these signals
to and from this modifed speech center? Then there's the nanotech piece -
simply binding these nervous system enhancements to a mechanical

Anyway, I'm not saying "this should be a simple project, let's have a
prototype ready in 6 months", I'm asking "what do you think about this
idea?". I don't deserve a critical lecture... Besides, are we focused only
on items which will be achievable within our lifetimes?

| > Education equates to software development... Implications?
| As The Great Cathulhu pointed out, the similarities are small. I would
| actually say practically zero. Software development has a definite
| goal, which is sought using various methodologies building a
| system. Education has a very loose goal of creating a great adult
| human being, but is an interaction between a learning agent and an
| environment where other agents try to achieve this goal.

Earlier I also clarified my supposition on sw vs. education - the gist:
software development is to a machine what learning is to a human, in a very
specific sense. When programming a machine, you're defining specific
reactions to stimuli (events). Experience/learning have a similiar effect
on a human child - given, the environment is more... anomalous. Consider
though, the process of teaching a child to read. First, constants are
clearly defined (alphabet). Next, a structure is developed for storing
groupings of these constants (words) - sound anything like a database
schema? Then, rules (logic) are developed which determine how words (and
series of words - sentences) should be interpreted... As you implied, there
is probably a difference between teaching a set of rules to a child, and
teaching them "how to think" - I'm only suggestiong the similiarity of the
former to software development.

| This goes both for todays humans and for augmented humans tomorrow -
| if I had a chip in my head providing instant access to information
| when I was a small kid, it would still just have been a high-potence
| information source I had to integrate into my growing cognitive
| network. If an augmented human can be designed like a software
| project, then either the entities doing the project are of such a high
| order of capability that they are definitely posthuman, or the
| "augmented" human will be extremely simple and predictable. True
| learning and education is a synthesis of various forms of information
| into new and useful cognitive patterns. This is unpredictable and
| highly individual - even simple neural networks develop different
| internal representations when you train them in different runs.

I *completely* agree - I am not suggesting we apply software methodologies
to human development, quite the opposite. We would simply be "training" any
mentally enhanced structures using the same tool that is currently used for
other aspects of mental development - learning.

| I think gardening is the best metaphor I can come up with for
| education right now.
| --
| -----------------------------------------------------------------------
| Anders Sandberg Towards Ascension!
| GCS/M/S/O d++ -p+ c++++ !l u+ e++ m++ s+/+ n--- h+/* f+ g+ w++ t+ r+ !y

This archive was generated by hypermail 2b29 : Thu Jul 27 2000 - 14:11:10 MDT