[ExI] Some new angle about AI.

Aware aware at awareresearch.com
Thu Jan 7 17:54:34 UTC 2010


2010/1/7 John Clark <jonkc at bellsouth.net>:
> On Jan 6, 2010, at 1:20 PM, x at extropica.org wrote:
> Me:
>
>>> we learned from the history of Evolution that consciousness is easy but
>>> intelligence is hard.
>
>> So why don't you agree with me that intelligence must have "existed"
>> (been recognizable, if there had been an observer) for quite a long
>> time
>
> Because we learned from the history of Evolution that consciousness is easy
> but
> intelligence is hard.

Well, that response clearly adds nothing to the discussion, and you
stripped out my supporting text.


>> before evolutionary processes stumbled upon the additional,
>> supervisory, hack of self-awareness
>
> What you just said is logically absurd.

Really?  Given your experience of my thinking over several years
together on this list, do you think it's more likely that I'm simply a
generator of logical absurdities, or is it more likely that you don't
understand the basis of my statement?  [I note that you're not asking
for any clarification.]


> If consciousness doesn't effect intelligence

Do you mean literally "If consciousness doesn't produce intelligence"
or do you mean "If consciousness doesn't affect intelligence"?

If you mean literally the former, then it appears that you must harbor
a mystical notion of "consciousness", that contributes to the somewhat
"intelligent" behavior of the amoeba, despite its apparent lack of the
neuronal apparatus necessary to support a sense of self.  I know John
Clark doesn't tolerate mysticism, and I know Damien has already your
use of the word "effect", so I can only guess that you mean
"consciousness" in a way that doesn't require much, if any, hardware
support.

[I'll note here that in addition to stripping out substantial portions
of my supporting text, you've also eliminated my careful definitions
of what I meant when I used the words "consciousness" and
"intelligence."]


> then there is no way Evolution could have "stumbled upon" the
> trick of generating consciousness

It may be relevant that the way evolution (It's not clear why you
would capitalize that word) works is always in terms of blind,
stumbling, random variation.  Of course genetic variation is strongly
constrained, and phenotypic variation is strongly facilitated by
preexisting structures.

> because it would convey no more adaptive
> advantage than eyes or pigment does for creatures that live all their life
> in dark caves.

There appears to be such a strong disconnect here that I suspect we're
not even talking about the same things.  It seems obvious that given a
particular degree of adaptation of an organism to its environment,
then to the extent the organism's fitness would be enhanced by the
ability to model possible variations on itself acting within its
environment, especially if this facilitates cooperative behaviors with
others similar to itself, such "adaptive advantage" would tend to be
selected.  What do YOU mean?


> In short if even one conscious being exists on Planet Earth
> and if Evolution is true then the Turing Test works;

Huh?  If there were only one conscious being, then wouldn't that have
to be the one judging the Turing Test?  And if there is no other
conscious being, how could any (non-conscious by definition) subject
pass the test (such that the TT would be shown to "work"?


> and if the Turing Test
> doesn't work then neither does Evolution.

Huh??


>> novel hacks like self-awareness discovered at some point, exploited
>> for the additional fitness they confer
>
> Fine, if true and consciousness aids fitness then it can be deduced from
> behavior.

Well, not "deduced" but certainly inferred...

> Either you can have intelligence without consciousness or you can
> not. The propositions

I'm going to assume, since you emphasize "Evolution", that your
propositions should be stated in terms of "evolved organisms", and not
in terms of the more general "systems that display behavior assessed
as intelligent."

So,
(A) Evolved organisms can be correctly assessed as displaying
intelligence but without consciousness.
(B) Evolved organisms can be correctly assessed as displaying
intelligence along with consciousness.

> lead to mutually contradictory conclusions, they can't
> both be right and you can't claim both as your own. You've got to make a
> choice.

Why? It seems to me that we observe the existence of both classes of
evolved organisms.  As I've said before, a wide range of organisms can
display behaviors expressing a significant degree of effective
prediction and control in regard to their environment of adaptation.
And the addition of a layer of supervisory self-awareness can be a
beneficial add-on for more advanced environments of interaction.

I'm guessing that our disagreement here comes down to different usage
and meaning of the terms "intelligence" and "consciousness" and it
might be significant that you stripped out all evidence and results of
my efforts to effectively define them.

You seem not to play fair, so it's not much fun.

- Jef



More information about the extropy-chat mailing list