Re: Can brains be good or evil?

On 4 Mar 2006 18:20:04 -0800, "JGCASEY" <jgkjcasey@xxxxxxxxxxxx> in
comp.ai.philosophy wrote:


Lester Zick wrote:
On 4 Mar 2006 13:42:10 -0800, "JGCASEY" <jgkjcasey@xxxxxxxxxxxx> in
comp.ai.philosophy wrote:


Lester Zick wrote:
On 4 Mar 2006 11:12:59 -0800, "feedbackdroids"

...
It doesn't work in a vacuum. Adding feedback to a
linear amplifier, or a feedback governor to a steam
engine, doesn't suddenly make them intelligent [as
in human-level]. It still takes 100B neurons,
operating properly and interacting properly, to get
intelligence. It's a system and has many components.
It's the manner in which FB is **used** that produces
intelligence.

So if adding feedback to linear amps or steam engines
doesn't make them intelligent what is this "properly
operating and interacting" property of 100B neurons
that "just" makes them intelligent?
...

I would say that feedback produces a simple form of
intelligent behavior in that it produces goal orientated
behavior rather than locking up or moving randomly
from one state to another.

So intelligence is goal oriented behavior? This may be
true but it isn't exhaustive. Negative feedback produces
a tendency toward stasis but lots of purely material
systems direct towards stasis.

Our goals may be a lot more complex than that of a
steam engine in that we have a lot more variables
that describe the goal and they change continually
but we still use feedback to move toward those goals
or stable states.

So intelligence is feedback with lots of variables?
I imagine there are lots of purely physical systems
directed towards various goals with lots of variables.

It is not so much that "intelligence" is this or
that but rather some of these things are required for
"intelligence" to occur.

Okay. But the question then becomes which are some of those things
and why are they required for intelligence in mechanical terms to the
exclusion of other things? I can certainly agree that feedback in some
form is required to organize intelligence just as the senses are. But
that doesn't make feedback or sensation necessarily intelligent.

Thus if a machine or animal shows goal orientated behavior
we say it is showing a degree of intelligence. It doesn't
mean it has to be "alive" or "conscious of what it is doing"
although some people add this as a requirement.

Well I'm not really sure what "goal oriented behaivor" actually means
in mechanical terms. I think we can see it a lot of the time but using
that as an exhaustive criterion for sentience or intelligence means we
need to understand "goal oriented behavior" in exhaustive mechanical
and not just intuitive terms. Otherwise we're not doing science.

Negative feedback leading to some stable state/s or a
complex system (lots of variables) do not in themselves
mean intelligence but may be a requirement for it.

Okay. I think we can all agree on "stable states". But I still don't
see "lots of variables" as any kind of definition for complexity in
terms of intelligence.

Just as a lot of memory and a high speed cpu may not
mean a viable visual system for a robot, but it may
be a requirement for it to happen.

I don't agree if by "high speed cpu" you mean TvN mechanics because
that would imply that intelligence is a computable number and that is
precisely the kind of ambiguity that has to be addressed and resolved
in mechanical terms for the definition of intelligence. TvN mechanics
is wonderful. It just isn't intelligent unless intelligence is a
computable number and no one has shown that to be the case. Robotics
is different. Here there is every reason to believe robotics involves
computable numbers exclusively.

What if I was to sit down and add numbers together without
any reason, purpose, goal would you consider that intelligent
or stupid?

Actually "to sit down and add numbers together without any reason,
purpose, goal" would be your goal. But I don't understand what this
has to do with any criterion for intelligence.

No matter what skill is required for a task we
assume a goal, a reason for its executuion for the behavior
to be deemed "intelligent".

So? Goals are mechanically ambiguous. We can speak of a goal in a
river's flowing to the sea. That's not the same kind of goal we mean
in terms of intelligence. What we need is an exhaustive criterion for
intelligence and not just a bunch of inexhaustive circumstances. I
agree feedback, goal orientation, etc. are necessary for organization
of intelligence but not as an exhaustive criterion in itelf.

But I think we can definitely say that if matter interacts solely in
terms of differences between various pieces of matter and doesn't
interact in the absence of differences, that the only possible
exhaustive definition for intelligence as distinct from material
interactions generally is differences between differences and that
principle compounded in terms of itself.

Then we can get away from all this mechanical ambiguity relating to
goals, feedback, stasis, language etc. because we can then talk about
intelligence as opposed to matter in mechanical terms and stop trying
to second guess ourselves about what each of those terms is supposed
to mean in this context or that context and whether rocks can talk or
not talk because we'll at last understand the difference between rocks
and intelligence to begin with and not have to ponder the motives or
goals of rocks as opposed to sentience and intelligence.

~v~~

.

Relevant Pages

  • Re: logical laws of Nature
    ... Of course the simple feedback used to in a crude guidance ... of humans the goals are much more complex and the means ... terminology doesn't help me. ... It is not the stuff I am calling purpose or intention it is ...
    (comp.ai.philosophy)
  • Re: Can brains be good or evil?
    ... engine, doesn't suddenly make them intelligent [as ... So if adding feedback to linear amps or steam engines ... So intelligence is goal oriented behavior? ... directed towards various goals with lots of variables. ...
    (comp.ai.philosophy)
  • Re: Website for Comments and or Suggestions
    ... One of my personal goals here in this newsgroup is to offer "constructive" ... website and ask for feedback, and I admire and respect that. ... Look, I am not an expert web designer, and my opinions are just that. ... Here are two links that might help in planning the design of your site: ...
    (microsoft.public.publisher.webdesign)
  • Re: Hoist V3 opinions?
    ... Thank you for all for your feedback. ... My training goals, as best as I can describe them (I bike about 10 KM ... I was consdering a home gym because of the range of exercises available ... no I was not looking for one affirmation but solid recommendations ...
    (misc.fitness.weights)
  • Re: Jeff Hawkins Q&A
    ... the impression that's exactly what you consider feedback. ... why not intelligence and consciousness. ... that's why I stress the mechanics of sentience so much. ... which people don't realize involve circular regressions. ...
    (comp.ai.philosophy)