Abram,
Thanks. V. helpful and interesting. Yes, on further examination, these
interactionist guys seem, as you say, to be trying to take into account the
embeddedness of the computer.
But no, there's still a huge divide between them and me. I would liken them
in the context of this discussion, to Pei who tries to argue that NARS is
"non-algorithmic", because the program is continuously changing. - and
therefore satisfies the objections of classical objectors to AI/AGI.
Well, both these guys and Pei are still v. much algorithmic in any
reasonable sense of the word - still following *structures,* if v.
sophisticated (and continuously changing) structures, of thought.
And what I am asserting is a paradigm of a creative machine, which starts
as, and is, NON-algorithmic and UNstructured in all its activities, albeit
that it acquires and creates a multitude of algorithms, or
routines/structures, for *parts* of those activities. For example, when you
write a post, nearly every word and a great many phrases and even odd
sentences, will be automatically, algorithmically produced. But the whole
post, and most paras will *not* be - and *could not* be.
A creative machine has infinite combinative potential. An algorithmic,
programmed machine has strictly limited combinativity..
And a keyboard is surely the near perfect symbol of infinite, unstructured
combinativity. It is being, and has been, used in endlessly creative ways -
and is, along with the blank page and pencil, the central tool of our
civilisation's creativity. Those randomly arranged letters - clearly
designed to be infinitely recombined - are the antithesis of a programmed
machine.
So however those guys account for that keyboard, I don't see them as in any
way accounting for it in my sense, or in its true, full usage. But thanks
for your comments. (Oh and I did understand re Bayes - I was and am still
arguing he isn't valid in many cases, period).
Mike,
The reason I decided that what you are arguing for is essentially an
interactive model is this quote:
"But that is obviously only the half of it.Computers are obviously
much more than that - and Turing machines. You just have to look at
them. It's staring you in the face. There's something they have that
Turing machines don't. See it? Terren?
They have - a keyboard."
A keyboard is precisely what the interaction theorists are trying to
account for! Plus the mouse, the ethernet port, et cetera.
Moreover, your general comments fit into the model if interpreted
judiciously. You make a distinction between rule-based and creative
behavior; rule-based behavior could be thought of as isolated
processing of input (receive input, process without interference,
output result) while creative behavior is behavior resulting from
continual interaction with and exploration of the external world. Your
concept of organisms as "organizers" only makes sense when I see it in
this light: a human organizes the environment by interaction with it,
while a Turing machine is unable to do this because it cannot
explore/experiment/discover.
-Abram
On Thu, Sep 4, 2008 at 1:07 PM, Mike Tintner <[EMAIL PROTECTED]>
wrote:
Abram,
Thanks for reply. But I don't understand what you see as the connection.
An
interaction machine from my brief googling is one which has physical
organs.
Any factory machine can be thought of as having organs. What I am trying
to
forge is a new paradigm of a creative, free machine as opposed to that
exemplified by most actual machines, which are rational, deterministic
machines. The latter can only engage in any task in set ways - and
therefore
engage and combine their organs in set combinations and sequences.
Creative
machines have a more or less infinite range of possible ways of going
about
things, and can combine their organs in a virtually infinite range of
combinations, (which gives them a slight advantage, adaptively :) ).
Organisms *are* creative machines; computers and robots *could* be (and
are,
when combined with humans), AGI's will *have* to be.
(To talk of creative machines, more specifically, as I did, as
keyboards/"organisers" is to focus on the mechanics of this infinite
combinativity of organs).
Interaction machines do not seem in any way then to entail what I'm
talking
about - "creative machines" - keyboards/ organisers - infinite
combinativity
- or the *creation,* as quite distinct from *following* of
programs/algorithms and routines..
Abram/MT:>> If you think it's all been said, please point me to the
philosophy of AI
that includes it.
I believe what you are suggesting is best understood as an interaction
machine.
General references:
http://www.cs.brown.edu/people/dqg/Papers/wurzburg.ps
http://www.cs.brown.edu/people/pw/papers/ficacm.ps
http://www.la-acm.org/Archives/laacm9912.html
The concept that seems most relevant to AI is the learning theory
provided by "inductive turing machines", but I cannot find a good
single reference for that. (I am not knowledgable on this subject, I
just have heard the idea before.)
--Abram
-------------------------------------------
agi
Archives: https://www.listbox.com/member/archive/303/=now
RSS Feed: https://www.listbox.com/member/archive/rss/303/
Modify Your Subscription: https://www.listbox.com/member/?&
Powered by Listbox: http://www.listbox.com
-------------------------------------------
agi
Archives: https://www.listbox.com/member/archive/303/=now
RSS Feed: https://www.listbox.com/member/archive/rss/303/
Modify Your Subscription:
https://www.listbox.com/member/?&
Powered by Listbox: http://www.listbox.com
-------------------------------------------
agi
Archives: https://www.listbox.com/member/archive/303/=now
RSS Feed: https://www.listbox.com/member/archive/rss/303/
Modify Your Subscription:
https://www.listbox.com/member/?&
Powered by Listbox: http://www.listbox.com
-------------------------------------------
agi
Archives: https://www.listbox.com/member/archive/303/=now
RSS Feed: https://www.listbox.com/member/archive/rss/303/
Modify Your Subscription:
https://www.listbox.com/member/?member_id=8660244&id_secret=111637683-c8fa51
Powered by Listbox: http://www.listbox.com