Re: How would a computer know if it were conscious?

From: David Nyman <david.nyman.domain.name.hidden>
Date: Sat, 23 Jun 2007 15:58:39 +0100

On 23/06/07, Russell Standish <lists.domain.name.hidden> wrote:

RS: I don't think I ever really found myself in
disagreement with you. Rather, what is happening is symptomatic of us
trying to reach across the divide of JP Snow's two cultures. You are
obviously comfortable with the world of literary criticism, and your
style of writing reflects this. The trouble is that to someone brought
up on a diet of scientific and technical writing, the literary paper
may as well be written in ancient greek. Gibberish doesn't mean
rubbish or nonsense, just unintelligible.

DN: It's interesting that you should perceive it in this way: I hadn't
thought about it like this, but I suspect you're not wrong. I haven't
consumed very much of your 'diet', and I have indeed read quite a lot of
stuff in the style you refer to, although I often find it rather
indigestible! But on the other hand, much of my professional experience has
been in the world of computer programming, right back to machine code days,
so I'm very aware of the difference between 'syntax' and 'semantics', and I
know too well how consequences can diverge wildly from a difference of a
single bit. How often have I heard the beleaguered self-tester wail "I
didn't *mean* that!"

So - to me - my process is a bit like: define a 'procedural language'; use
this to 'code' some 'problem'; 'run' it to see what happens; then 'debug'
and repeat. This is no doubt excruciating for anyone else to follow, and my
attempts to 'comment' the code don't always help. Now that I'm re-reading
TON, it seems to me that I've been trying to re-interpret bits of it in this
way (in an attempt to reconcile what you and Colin were disputing) but only
succeeded in muddying the waters further, probably for the reasons you
suggest.

However, in the spirit of the original topic of the thread, I would prefer
to ask you directly about the plausibility (which, unless I've
misunderstood, you support?) of an AI-program being in principle
'conscious'. I take this to entail that instantiating such a program
thereby implements an 'observer' that can respond to and share a reality, in
broadly the same terms, with human 'observers'. (I apologise in advance if
any paraphrase or short-hand I adopt misrepresents what you say in TON):

TON, as you comment in the book, takes the 'idealist' stance that 'concrete'
notions emerge from observation. Our own relative status as observers
participating in 'worlds' is then dependent on computational 'emergence'
from the plenitude of all possible bit-strings. Let's say that I'm such an
observer and I observe a 'computer' like the one I'm using now. The
'computer' is a 3-person 'concrete emergent' in my 1-person world, and that
of the 'plurality' of observers with whom I'm in relation: we can 'interact'
with it. Now, we collectively *impute* that some aspect of its 3-person
behaviour (e.g. EM phenomena in its internal circuitry) is to be regarded as
'running an AI program' (i.e. ISTM that this is what happens when we
'compile and run' a program). In what way does such imputation entail the
evocation - despite the myriad possible 'concrete' instantiations that might
represent it - of a *stable* observer capable of participating in our shared
'1-person plural' context? IOW, I'm concerned that two different categories
are being conflated here: the 'world' at the 'observer level' that includes
me and the computer, and the 'world' of the program, which is 'nested'
inside this. How can this 'nested' world get any purchase on 'observables'
that are 'external' to it?

As I re-read this question, I wonder whether I've already willy-nilly fallen
into the '2-cultures' gap again. But what I've asked seems to be directly
related to the issues raised by 'Olympia and Klara', and by the substitution
level dilemma posed by 'yes doctor'. Could you show me where - or if - I go
wrong, or does the 'language game' make our views forever mutually
unintelligible?

David

>
> On Fri, Jun 22, 2007 at 02:06:14PM +0100, David Nyman wrote:
> > RS:
> > Terminology is terminology, it doesn't have a point of view.
> >
> > DN:
> > This may be a nub of disagreement. I'd be interested if you could
> clarify.
> > My characterisation of a narrative as '3-person' is when (ISTM) that
> it's an
> > abstraction from, or projection of, some 'situation' that is
> fundamentally
> > 'participative'. Do you disagree with this?
> >
> > By contrast, I've been struggling recently with language that engages
> > directly with 'participation'. But this leads to your next point.....
>
> Terminology is about describing communicable notions. As such, the
> only things words can ever describe are 1st person plural
> things. Since you are familiar with my book, you can look up the
> distinction between 1st person (singular), 1st person plural and 3rd
> person, but these concepts have often been discussed on this list. I
> can use the term "Green" for instance, in a sentence to you, and we
> can be sure of its meaning when referring to shared experience of
> phenomena, however I can never communicate to you how green appears to
> me, so that you can compare it with your green qualia.
>
> >
> > RS:
> > Terms
> > should have accepted meaning, unless we agree on a different meaning
> > for the purposes of discussion.
> >
> > DN:
> > But where there is no generally accepted meaning, or a disputed one, how
> can
> > we then proceed? Hence my attempts at definition (which I hate BTW),
> and
> > which you find to be gibberish. Is there a way out of this?
> >
>
> This sometimes happens. We can point to examples of what the word
> means, and see if we agree on those. There are bound to be borderline
> cases where we disagree, but these are often unimportant unless we are
> searching for a definition.
>
> > BTW, when I read 'Theory of Nothing', which I find very cogent, ISTM
> that
> > virtually its entire focus is on aspects of a 'participatory'
> approach. So
> > I'm more puzzled than ever why we're in disagreement.
>
> You are correct that it is 'particpatory', at least in the sense John
> Wheeler uses it. I don't think I ever really found myself in
> disagreement with you. Rather, what is happening is symptomatic of us
> trying to reach across the divide of JP Snow's two cultures. You are
> obviously comfortable with the world of literary criticism, and your
> style of writing reflects this. The trouble is that to someone brought
> up on a diet of scientific and technical writing, the literary paper
> may as well be written in ancient greek. Gibberish doesn't mean
> rubbish or nonsense, just unintelligible.
>
> I had my first experience of the modern academic humanities just two
> years ago, and it was quite a shock. I attended a conference entitled
> "The two cultures: Reconsidering the division between the Sciences and
> Humanities". I was invited to speak as one of the scientific
> representatives, and basically spoke about the core thesis of my book,
> which seemed appropriate. I kept the language simple and
> interdisciplinary, used lots of pictures to illustrate the concepts,
> and I'm sure had a reasonable connect with the audience. All of the
> other scientists did the same. They all knew better than to fall back
> into jargon and dense forests of mathematical formulae (I have suffered
> enough of those types of seminars, to be sure).
>
> By contrast, the speakers from the humanities all read their papers
> word-for-word. There were no illustrations to help one follow the gist
> of the arguments. The sentences were long-winded, and attempted to
> cover every nuance possible. A style I'm sure you're very familiar
> with. I tried to ask a few questions of the speakers at the end, not
> so as to appear smart or anything, but just to try to clarify some of
> the few points I thought I might have understood. The responses from
> the speakers, however, was in the same long-winded, heavily nuanced
> sentences.
>
> The one thing I drew from this conference was that the divide between
> Snow's two cultures is alive and well, and vaster than I ever imagined.
>
>
> > I've really been
> > trying to say that points-of-view (or 'worlds') emerge from *structure*
> > defined somehow, and that (tautologically, surely) the 'primitives' of
> such
> > structure (in whatever theoretical terms we choose) must be capable of
> > 'animating' such povs or worlds. IOW povs are always 'takes' on the
> whole
> > situation, not inherent in individuated 'things'.
>
> To say that a "point of view" (which I would translate as "observer")
> emerges from the worlds structure, is another way of saying that the
> observer must supervene on observed physical structures. And I agree
> with you, basically because of the "Occam catastrophe"
> problem. However, how or why this emergence happens is rather
> mysterious.
>
> I think is has something to do with self-awareness, without a self
> existing with the observed physical world, one cannot be
> self-aware. The corrolary of this is that self-awareness must be
> necessary for consciousness. Note this doesn't mean that you have to
> be self-aware every second you are awake, but you have to be capable
> of introspection.
>
> >
> > RS:
> > 2) Oxygen and hydrogen atoms as counterexamples of a chemical
> > potential that is not an electric field
> >
> > DN:
> > I certainly didn't mean to imply this! I just meant that we seemed to
> be
> > counterposing 'abstracted' and 'participative' accounts, in the sense I
> > indicate above. Something would really help me at this point: could I
> ask
> > how would you relate 'physical' levels of description you've used (e.g.
> > 'oxygen and hydrogen atoms') to the 'participative' approach of 'TON'?
> > IOW, how do these narratives converge on the range of phenomena to be
> > explained?
> >
> > David
> >
>
> I don't see a problem with this. Some sentences are necessarily
> reflective "Joan gave the ball to me", and other aren't "Joan gave the
> ball to Russell". A lot of phenomena can be adequately described
> without reference to any observer. Others require the presence of the
> observer to be explicitly acknowledged.
>
> --
>
>
> ----------------------------------------------------------------------------
> A/Prof Russell Standish Phone 0425 253119 (mobile)
> Mathematics
> UNSW SYDNEY 2052 hpcoder.domain.name.hidden
> Australia http://www.hpcoders.com.au
>
> ----------------------------------------------------------------------------
>
> >
>

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "Everything List" group.
To post to this group, send email to everything-list.domain.name.hidden
To unsubscribe from this group, send email to everything-list-unsubscribe.domain.name.hidden
For more options, visit this group at http://groups.google.com/group/everything-list?hl=en
-~----------~----~----~----~------~----~------~--~---
Received on Sat Jun 23 2007 - 11:00:41 PDT

This archive was generated by hypermail 2.3.0 : Fri Feb 16 2018 - 13:20:14 PST