Newsgroups: comp.ai.philosophy
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!europa.eng.gtefsd.com!howland.reston.ans.net!pipex!uknet!dcs.gla.ac.uk!unix.brighton.ac.uk!mjs14
From: mjs14@unix.brighton.ac.uk (shute)
Subject: Re: Minsky's new article (was: Roger Penrose's new book)
Message-ID: <1994Nov7.142110.22111@unix.brighton.ac.uk>
Organization: University of Brighton, UK
References: <39al9e$beq@coli-gate.coli.uni-sb.de> <39bg6v$9if@cantaloupe.srv.cs.cmu.edu> <39d8g2$dlm@coli-gate.coli.uni-sb.de>
Date: Mon, 7 Nov 1994 14:21:10 GMT
Lines: 58

In article <39d8g2$dlm@coli-gate.coli.uni-sb.de> sean@mpi-sb.mpg.de (Sean Matthews) writes:
>Now it may very well be that in the
>future [...] we are all going to turn into super
>powerful robots with brains the size of planets, and live for ever.
>There is no *a priori* reason that I know of why this should not
>happen, but Minsky provides no grounds for believing that it will or
>even can happen, beyond that he clearly thinks it would be wonderful
>if it did (a variant on a common wish fulfilment fantasy).

When someone mentioned the October issue of SA in this group, I rushed out
to buy it (luckily, the 3 week delay that we incur on this side of the pond
*sometimes* works in our favour as far as advanced warning of publications
is concerned :-).  I read each article with relish... learning much from
the cosmology and evolution articles.

Then I came to Marvin Minksky's article... and, like you, my initial reaction
was one of disappointment.  BUT... the more I see the article discussed,
the more I realise I got out of it.

As someone else pointed out, he was addressing a lay audience, so might not
have been able to accelerate them up the philosophy learning curve better
by any other route.  I could add to that, that he was also addressing an
informed audience who might yawn if he had taken the route that is traditional
in this newsgroup (building an artificial conscience from scratch).

I'd never thought of approaching the argument from this direction... and
the article has shed quite a new perspective on A.C., to my point-to-view.

In the light of this... what *I* would now envisage happening would be
artificial replacements (silicon, or whatever) for *parts* of brain
(either damaged by near fatal accident, or by serious mental deficiency)
similar to present-day heart/lung/liver/etc transplants.  Patients who
feel (or whose relations feel) that they have little life to lose, and
perhaps much to gain, by volunteering as surgical guinea-pigs.

Then, gradually, as lessons are learned, the surgery can: (a) become
established, and offered more routinely; and (b) the experimental side
can be made increasely more adventurous, to be offered to the next generation
of guinea-pigs.

By (a) we'd see the rich and famous able to choose to have new functionality
stitched on to their brains (the idea of a digital clock, or calculator
chip, addressable from the cortex, say);  by (b) we'd see more and more
of the natural neural process being taken over by silicon (or whatever) to
replace defective tissue.


So, in summary, despite my immediate reaction to Marvin Minsky's article,
I think that he will probably be right... but before his vision of robot
brains completely replacing human ones, there will be many
intermediate stages along the way.

(All of this, of course, would have to be reviewed if the dualists turn
out to have other future cards to play :-)
-- 

Malcolm SHUTE.         (The AM Mollusc:   v_@_ )        Disclaimer: all

