| View previous topic :: View next topic |
| Author |
Message |
Wolf Kirchmeir
Guest
|
 Posted: Wed Nov 19, 2008 9:52 pm Post subject: Re: A Critique of Prof. Hubert Dreyfus' "Why Heideggerian AI |
 |
|
Alpha wrote:
| Quote: | On Nov 17, 5:01 pm, c...@kcwc.com (Curt Welch) wrote:
[...]
All you can do from within philosophy is create multiple possible answers.
Incorrect; philosophy, in relation to science, is the *question-
generating* or *question-vetting* operation.
[...] |
Yes, that's what philosophers claim, and when it's done right (eg, tests
the logic of the arguments, drilling down to the underlying assumptions,
etc), then it can point up conceptual glitches, which in turn may lead
to useful reformulations the question(s).
But scientists are as good at doing this as are professional
philosophers. Consider the recent attempts to remodel the history of the
universe to eliminate the physically impossible singularity at the
moment of the Big Bang. If space is assumed to be granular (ie, that
space is not a continuum, but that there smallest bits of space), then a
theory that eliminates that singularity is possible. But that assumption
in turn raise the question of what those smallest bits of space are
"embedded in." Which is a metaphor arising from our direct experience of
objects in space.
How can there be smallest bits of something without those smallest bits
being in something else? IOW, how can there be objects without some
space for those objects to inhabit? That's a philosophical question. My
metaphysics allows for it, because I think that "If there is an action,
there must be actor" is fallacious. IOW, "There can be action with no
actor." Analogously, I can accept "There can be objects with no space."
If that sounds muddled, no surprise. It's damn difficult to avoid
thinking in the patterns that language embodies. It's damn difficult to
think in ways that run counter to what language allows as
sensible/sense-making utterances.
"In mathematics, we can know whether what we say is true, but we cannot
know what we are talking about. In poetry, we can know what we are
talking about, but we cannot know whether what we are saying is true."
(after Bertrand Russell.)
"Philosophers are poets who attempt logical proof of their metaphors."
(Wolf Kirchmeir -- you read it here first  .)
HTH
--
Wolf Kirchmeir |
|
| |
|
 Back to top |
Alpha
Guest
|
| Posted: Wed Nov 19, 2008 11:39 pm Post subject: Re: Critique #1 Re: A Critique of Prof. Hubert Dreyfus' "Why |
|
|
On Nov 19, 10:44 am, Publius <m.publ...@nospam.comcast.net> wrote:
| Quote: | forbisga...@msn.com wrote innews:57ff8b1b-67de-4784-94a1-a57f5bd3be0b@z6g2000pre.googlegroups.com:
It is more likely an emergent property of a particular implementation
strategy for intelligence.
There used to be a notion that an eye couldn't develop because of
its complexity. The assertion was that no where along the way
would the components have any survival value. I believe that dead
horse has been whipped enoungh. It seems to me that the same
applies to the development of consciousness in our set of evolved
systems. I'm not completely sure what the precursors were.
Quite a bit of work has been done that bears on that question, e.g.,
Metzinger, Edelman & Tononi, et al. The various sensory subsystems evolve
"N-maps" of the world as it is presented via their particular input
channels. These mapping schemes may all follow a common template, or
pattern, derived from some earlier integrative algorithm. Then the brain
"maps itself" --- it integrates all the N-maps into a comprehensive,
dynamic "world model" which represents the organism itself situated in its
environment, with each N-map contributing a dimension to this n-dimensional
model. The system thereafter interacts with the world "through the model,"
and the model is revised as the organism's actions affect it.
Included in the model are not only "state data" --- the current values of
the many input variables (colors, shapes, arrays, odors, etc. currently or
most recently perceived), but also the "transformation rules" of the
system --- under what conditions a given state is replaced by another
state. The system learns those rules through ongoing observation and
interaction with the world. That gives us a picture of the geometry and
dynamic of the external environment, its "laws of motion," which are then
incorporated into the model.
The great advantage of the dynamic world model is that it allows the system
to "run scenarios," i.e., to pose "what if?" questions, run the model with
the hypothetical data, and thus anticipate the outcomes of possible
actions. Such a system has a great survival advantage over systems which
can only respond to environmental changes in real time, no matter how
sophisticated their processing algorithms.
We can probably say that any system which can dynamically model itself, its
environment, and its ongoing interactions with the environment, is
conscious.
As with the eye, certain functional systems can be designed
without passing through all the evolutionary precursor steps.
This still leaves the open question as to whether or not
the implementation of the function(s) from which human
consciousness emerges necessarily entails that consciousness
will emerge from all implementations of the function(s).
I suspect any self-modeling system interacting with a dynamic environment
will be conscious.
|
Your suspicions are misplaced. Supose that an auto has a model of it
"self" residing in a hard drive as part of the auto. The model can
include anything you wish WRT the auto - operation processes, the
physics of combustion engines - a complete ontology of the auto from
all aspects. The result will not be a conscious auto as if by magic
or any other means.
| Quote: | Consciousness is a kind of "byproduct" of a particular
strategy for improving intelligence.
|
No, it is not a byproduct. It arises as a result of a living
autopoietic system that interacts with an environment (so far as we
know only living things have consciousness), independent of the
intelligence of the system/lifeform. As such it is not existent
*because* of a strategy for improving intelligence; rather, it does
serve to generate feedback to processes that might so operate to
increase (or decrease!) intelligence.
| Quote: | But intelligence (the capacity of a
system to solve novel problems) may be achieved in other ways.
|
That much may be true; Deep Blue solves chess problems without any
consciousness. |
|
| |
|
| Back to top |
Publius
Guest
|
| Posted: Wed Nov 19, 2008 11:44 pm Post subject: Re: Critique #1 Re: A Critique of Prof. Hubert Dreyfus' "Why |
|
|
forbisgaryg@msn.com wrote in
news:57ff8b1b-67de-4784-94a1-a57f5bd3be0b@z6g2000pre.googlegroups.com:
| Quote: | It is more likely an emergent property of a particular implementation
strategy for intelligence.
There used to be a notion that an eye couldn't develop because of
its complexity. The assertion was that no where along the way
would the components have any survival value. I believe that dead
horse has been whipped enoungh. It seems to me that the same
applies to the development of consciousness in our set of evolved
systems. I'm not completely sure what the precursors were.
|
Quite a bit of work has been done that bears on that question, e.g.,
Metzinger, Edelman & Tononi, et al. The various sensory subsystems evolve
"N-maps" of the world as it is presented via their particular input
channels. These mapping schemes may all follow a common template, or
pattern, derived from some earlier integrative algorithm. Then the brain
"maps itself" --- it integrates all the N-maps into a comprehensive,
dynamic "world model" which represents the organism itself situated in its
environment, with each N-map contributing a dimension to this n-dimensional
model. The system thereafter interacts with the world "through the model,"
and the model is revised as the organism's actions affect it.
Included in the model are not only "state data" --- the current values of
the many input variables (colors, shapes, arrays, odors, etc. currently or
most recently perceived), but also the "transformation rules" of the
system --- under what conditions a given state is replaced by another
state. The system learns those rules through ongoing observation and
interaction with the world. That gives us a picture of the geometry and
dynamic of the external environment, its "laws of motion," which are then
incorporated into the model.
The great advantage of the dynamic world model is that it allows the system
to "run scenarios," i.e., to pose "what if?" questions, run the model with
the hypothetical data, and thus anticipate the outcomes of possible
actions. Such a system has a great survival advantage over systems which
can only respond to environmental changes in real time, no matter how
sophisticated their processing algorithms.
We can probably say that any system which can dynamically model itself, its
environment, and its ongoing interactions with the environment, is
conscious.
| Quote: | As with the eye, certain functional systems can be designed
without passing through all the evolutionary precursor steps.
This still leaves the open question as to whether or not
the implementation of the function(s) from which human
consciousness emerges necessarily entails that consciousness
will emerge from all implementations of the function(s).
|
I suspect any self-modeling system interacting with a dynamic environment
will be conscious. Consciousness is a kind of "byproduct" of a particular
strategy for improving intelligence. But intelligence (the capacity of a
system to solve novel problems) may be achieved in other ways. |
|
| |
|
| Back to top |
Neil W Rickert
Guest
|
| Posted: Thu Nov 20, 2008 5:21 am Post subject: Re: Critique #2 Re: A Critique of Prof. Hubert Dreyfus' "Why |
|
|
Publius <m.publius@nospam.comcast.net> writes:
| Quote: | Neil W Rickert wrote:
I am skeptical that there is much being retained in the form of
stored representations as "past data". It seems more likely that
the brain is a bit like a finely tuned instrument. The past data
has played a role in adjusting the tuning, but has not been retained.
That is a surprising claim. How do you account for the ability of the
system to draw a face or hum a tune "from memory"?
|
To continue with the "finely tuned instrument" metaphor, a finely
tuned instrument will resonate corresponding to its tuning. The idea
is that finely tuned neural circuits will resonate to the type of
information that was used in tuning them. Thus remembering is a
process of reconstructing and testing what is reconstructed to see
how well it resonates.
| Quote: | The "motor plan" to grab an object is likely quite crude, and
precise behavior results not from having a precise plan, but from
measuring performance during the motor action and adjusting it where
the measurement indicates it is off. This would make measurement
more important than computation.
More likely both are in play (not unlike guiding a spacecraft to Mars).
|
I'll agree with that.
| Quote: | Isn't the frame problem mostly (if not all) about filtering the
intractable sensory information of any situation into a Gestalt of
only meaningful, important information. This seems to be along the
lines of "common sense", which you cannot just ingore and expect to
meet or exceed human intelligence or behavior skills.
That's not the way the frame problem is usually described.
See http://en.wikipedia.org/wiki/Frame_problem for a more familiar
version.
The first sentence of the article exemplies its thesis: "In artificial
intelligence, the frame problem was initially formulated as the problem
of expressing a dynamical domain in logic without explicitly specifying
which conditions are not affected by an action."
With the phrase "in artificial intelligence", the writer frames the
scope of the inquiry and thus limits it.
|
But it was about limiting the number of logic forms, not about
limiting the sensory information.
| Quote: | In any case, what matters here is the version of the frame
problem being assumed by Dreyfus in the paper we are discussing.
And I'm pretty sure he is taking it as the problem of updating
stored representations so that they are consistent with changes in
the world. Thus he mentions that Rodney Brooks avoids the problem
by designing his system to not depend on stored representations of
the state of the world.
The implication of that thesis is that all problems are approached *de
novo*, and all useful or necessary information gathered only after the
problem presents itself.
|
In the case of Brooks work, that might be true. It is a criticism
of Brooks systems, that they don't learn. However, your conclusion
is not a necessary consequence of the assumption that stored
representations are not used. Earlier experience can be used to adjust
the response of the system, so that it will performs better in future.
This does not require the use of stored representations. |
|
| |
|
| Back to top |
Publius
Guest
|
| Posted: Thu Nov 20, 2008 8:02 am Post subject: Re: Critique #2 Re: A Critique of Prof. Hubert Dreyfus' "Why |
|
|
Neil W Rickert <rickert+nn@cs.niu.edu> wrote in news:pc1Vk.7456$x%.3332
@nlpi070.nbdc.sbc.com:
| Quote: | In the case of Brooks work, that might be true. It is a criticism
of Brooks systems, that they don't learn. However, your conclusion
is not a necessary consequence of the assumption that stored
representations are not used. Earlier experience can be used to adjust
the response of the system, so that it will performs better in future.
This does not require the use of stored representations.
|
Requires a more specific definition of "intelligence." Learning, skill
acquisition and refinement, i.e., operant conditioning, are not
"intelligence," or at least, not the most interesting aspect of it. I like
the definition, "The capacity of a system to solve novel problems."
How quickly can the system solve a problem it has never before encountered?
How does it manage to do it? Developing efficiency in dealing with familiar
or recurring problems is fairly straightforward. |
|
| |
|
| Back to top |
Neil W Rickert
Guest
|
| Posted: Thu Nov 20, 2008 9:07 am Post subject: Re: Critique #2 Re: A Critique of Prof. Hubert Dreyfus' "Why |
|
|
Publius <m.publius@nospam.comcast.net> writes:
| Quote: | Neil W Rickert <rickert+nn@cs.niu.edu> wrote in news:pc1Vk.7456$x%.3332
@nlpi070.nbdc.sbc.com:
In the case of Brooks work, that might be true. It is a criticism
of Brooks systems, that they don't learn. However, your conclusion
is not a necessary consequence of the assumption that stored
representations are not used. Earlier experience can be used to adjust
the response of the system, so that it will performs better in future.
This does not require the use of stored representations.
Requires a more specific definition of "intelligence." Learning, skill
acquisition and refinement, i.e., operant conditioning, are not
"intelligence," or at least, not the most interesting aspect of it. I like
the definition, "The capacity of a system to solve novel problems."
|
I never actually mentioned operant conditioning. I am not assuming
that learning is operant conditioning. No I am not assuming what
you take to be an uninteresting definition of intelligence.
We just disagree on how learning works.
| Quote: | How quickly can the system solve a problem it has never before encountered?
|
That question is too vague to be answerable.
| Quote: | How does it manage to do it? Developing efficiency in dealing with familiar
or recurring problems is fairly straightforward.
|
How about developing new abilities - again, this need not require
storing representations. |
|
| |
|
| Back to top |
Publius
Guest
|
| Posted: Thu Nov 20, 2008 9:07 am Post subject: Re: Critique #1 Re: A Critique of Prof. Hubert Dreyfus' "Why |
|
|
Alpha <omegazero2003@yahoo.com> wrote in
news:f4e53f60-3d2f-48c1-b489-4635bbf965be@w1g2000prk.googlegroups.com:
| Quote: | I suspect any self-modeling system interacting with a dynamic
environment will be conscious.
Your suspicions are misplaced. Supose that an auto has a model of it
"self" residing in a hard drive as part of the auto. The model can
include anything you wish WRT the auto - operation processes, the
physics of combustion engines - a complete ontology of the auto from
all aspects. The result will not be a conscious auto as if by magic
or any other means.
|
The model represents the system (insofar as it can represent itself using
data gathered through its own sensory apparatus), situated in an
environment. The model is constantly updated as the ongoing interaction
modifies both the system and the environment.
How would decide whether the system was or was not conscious?
| Quote: | Consciousness is a kind of "byproduct" of a particular
strategy for improving intelligence.
No, it is not a byproduct. It arises as a result of a living
autopoietic system that interacts with an environment (so far as we
know only living things have consciousness), independent of the
intelligence of the system/lifeform.
|
That will describe all living cells and organisms, not all of which (or
even most of which) are conscious. Unless you have adopted a definition of
"consciousness" so broad as to be meaningless.
It is true that only living systems exhibit consciousness --- so far. But
until 200 years or so ago, only living systems exhibited locomotion, and
until 50 years ago only living systems exhibited intelligence. There is no
reason to suppose consciousness requires a biological substrate. |
|
| |
|
| Back to top |
Publius
Guest
|
| Posted: Thu Nov 20, 2008 9:07 am Post subject: Re: Critique #1 Re: A Critique of Prof. Hubert Dreyfus' "Why |
|
|
curt@kcwc.com (Curt Welch) wrote in
news:20081119010159.274$s6@newsreader.com:
Interesting comments, and generally on the right track, but I have some
quibbles.
| Quote: | Well, I think people in general are very confused about the entire
subject of consciousness, so when they talk about it, they are often
talking about many different aspects of humans. However, I believe
the foundation of where all this confusion comes from, is in how the
brain tends to model itself. That is, our entire understanding and
perception of reality is created by the processing that happens in our
brain, and that includes our perception of our own brain. Our
perception of self.
|
There is no perception of self, except insofar as the "self" includes the
body. The "self" is a construct, a model of the system synthesized and
inferred from the current states of other brain subsystems (those which
process sensory data) and stored information regarding past states of those
subsystems. So we have a *conception* of self, not a perception.
| Quote: | Because the brain has only limited access to
itself, the model it creates is also limited. Based on the sensory
data the brain has access to, it is forced to model internally
generated brain signals as having no association with external sensory
signals.
|
Well, we do assume there is an association, actually. We also construct a
"world model," with the self-model situated within it. We accept this world
model as "the world" (we are all realists by default). Yet, because the
model remains available even when the world "goes away" (when we close our
eyes, change location, or just direct attention elsewhere), we conclude
there is a another "realm" where the world continues to exist --- "it
exists in the mind." The notion of "mind" arises because we are able to
contemplate aspects of the world not currently present to the senses,
including past states of the world.
| Quote: | The result of this fact is that the brain builds a model of this data
by indicating no association between private thoughts, and the
physical events represented in the sensory data. The result of that
is very simple - it leaves us with a model of realty where thoughts
are disconnected from all things physical.
|
We realize they are connected, but are nonetheless distinct.
But I agree with what I take to be your central point --- we are aware of
the model, but not of the brain mechanisms which generate it. So it becomes
conceptually detached from its substrate. |
|
| |
|
| Back to top |
Alpha
Guest
|
| Posted: Thu Nov 20, 2008 4:22 pm Post subject: Re: Critique #1 Re: A Critique of Prof. Hubert Dreyfus' "Why |
|
|
On Nov 19, 11:43 pm, Publius <m.publ...@nospam.comcast.net> wrote:
| Quote: | Alpha <omegazero2...@yahoo.com> wrote innews:f4e53f60-3d2f-48c1-b489-4635bbf965be@w1g2000prk.googlegroups.com:
I suspect any self-modeling system interacting with a dynamic
environment will be conscious.
Your suspicions are misplaced. Supose that an auto has a model of it
"self" residing in a hard drive as part of the auto. The model can
include anything you wish WRT the auto - operation processes, the
physics of combustion engines - a complete ontology of the auto from
all aspects. The result will not be a conscious auto as if by magic
or any other means.
The model represents the system (insofar as it can represent itself using
data gathered through its own sensory apparatus), situated in an
environment. The model is constantly updated as the ongoing interaction
modifies both the system and the environment.
|
Sure - but a toaster with some memory (model) of how much it burned
the last bread inserted (its effect on the "environment") would
qualify, and we know it isn't the least bit conscious.
| Quote: |
How would decide whether the system was or was not conscious?
|
That is of course the million dollar question! ;^) I suggest a
combination of factors including: architecture of the entity/
complexity, whether the system is autopoietic thence giving rise to an
emergent "self" that is reflected in behavior - reactions to
environmental perturbations, whethr it is living or dead (as we know
that dead things do not exhibit even a little bit of consciousness),
and so forth.
| Quote: |
Consciousness is a kind of "byproduct" of a particular
strategy for improving intelligence.
No, it is not a byproduct. It arises as a result of a living
autopoietic system that interacts with an environment (so far as we
know only living things have consciousness), independent of the
intelligence of the system/lifeform.
That will describe all living cells and organisms, not all of which (or
even most of which) are conscious. Unless you have adopted a definition of
"consciousness" so broad as to be meaningless.
|
It is not meaningless to suppose that living things have some measure
of consciousness, even if a little bit in the cases of bateria etc.;
for example, a following of a sugar gradient by a baterium to ensure
survival would indicate that there is some sense of individuality, as
set against an environment, is there. The enactive approach of
Thompson & Varela comes into play here.
| Quote: |
It is true that only living systems exhibit consciousness --- so far. But
until 200 years or so ago, only living systems exhibited locomotion, and
until 50 years ago only living systems exhibited intelligence. There is no
reason to suppose consciousness requires a biological substrate.
|
Except that such is the only substrate that even comes close to
exhibiting that faculty and there is no reason to believe that other
substrates could qualify considering what we have so far as artifacts
that are not living. Deep Blue is extremly good at modelling its own
state (an intelligence aspect) and the state of the playing scenario,
but is not the leat bit conscious for example. Neither is Curt's
keyboard! ;^) |
|
| |
|
| Back to top |
Josip Almasi
Guest
|
| Posted: Thu Nov 20, 2008 5:23 pm Post subject: Re: Critique #1 Re: A Critique of Prof. Hubert Dreyfus' "Why |
|
|
Isaac wrote:
| Quote: | "Josip Almasi" <joe@vrspace.org> wrote in message
Actually there's yet another big question - is consciousness emergent
property of (sufficiently high) intelligence.
I tend to disagree. I do not see the two being so intimately connected so
as to require one to immerge from another. Intelligence might just be a
boundary condition on the scope of consciousness (i.e., awareness). For
example, a severely retarded person is certainly far less intelligent,
however, I don't think there is any evidence that they are far less
conscious. If consciouness emerged (necessarily?) from intelligence then
shouldn't they be highly correlated?
|
Fair enough.
Then again, we must not forget social interactions.
Retards get much more attention and training than healthy kids.
Say, I've spent endless hours with a grown up retard (immaturitas
emotionalis, debilitas) unable to remember 3 word sequences. Takes 2
hours for a sequence, 2 hours later he forgets, then all over again;
eventually, he makes it.
There's schools for such people, so society makes up for their disabilities.
Regards... |
|
| |
|
| Back to top |
Josip Almasi
Guest
|
| Posted: Thu Nov 20, 2008 5:45 pm Post subject: Re: Critique #1 Re: A Critique of Prof. Hubert Dreyfus' "Why |
|
|
Publius wrote:
| Quote: | Josip Almasi <joe@vrspace.org> wrote in
news:gfujgq$4el$1@gregory.bnet.hr:
Actually there's yet another big question - is consciousness emergent
property of (sufficiently high) intelligence.
It is more likely an emergent property of a particular implementation
strategy for intelligence.
|
Well said.
| Quote: | As for empirical part, representationalism is sort of top-down
approach, while say Curt seems to prefer bottom-up approach.
Can't say for Dreyfus though:)
How natural intelligent systems evolved is a different question from their
structure and function. Artificial systems might be designed which have
different structures, but perform the same functions and are just as
intelligent (or more so).
|
I don't think it's general evolution of intelligence that matters here,
it's evoution of *a* system - can we give it representations or it has
to make them for itself.
Talking about *a* system, we might learn much from how kids learn. And
there, representationalists have big arguments from Chomsky & co, who
claim we're hardwired for language, IOW we get representations from our
earliest days.
Now that I said that, I think Dreyfus is offtopic with Heidegger:>
And WRT functions:) IIRC Ben Goertzel said once, my dog is sure more
intelligent for finding poop in the field, so what?;)
Regards... |
|
| |
|
| Back to top |
Josip Almasi
Guest
|
| Posted: Thu Nov 20, 2008 6:41 pm Post subject: Re: Critique #1 Re: A Critique of Prof. Hubert Dreyfus' "Why |
|
|
Curt Welch wrote:
| Quote: |
Well, I think people in general are very confused about the entire subject
of consciousness, so when they talk about it, they are often talking about
many different aspects of humans.
|
Agreed.
| Quote: | However, I believe the foundation of
where all this confusion comes from, is in how the brain tends to model
itself. That is, our entire understanding and perception of reality is
created by the processing that happens in our brain, and that includes our
perception of our own brain. Our perception of self. Because the brain
has only limited access to itself, the model it creates is also limited.
Based on the sensory data the brain has access to, it is forced to model
internally generated brain signals as having no association with external
sensory signals. It does this simply because there is no temporal
correlation between auditor sensory data and the brain signals which
represent our private thoughts. That is, we can not hear the brain making
any physical sound as our neurons fire. Because the brain makes no noise
(that our ears can pick up) when it operates, there is no correlation
between auditory sensory signals, and private thought signals. And because
there is no correlation, the brain doesn't create an association between
these different signals. This is true for all the physical sensory data
that flows into the brain. We can't hear our private thoughts, we can not
feel the brain vibrate, we can not smell it. There simply are no
correlations between our physical sensory data and our internal thought
signals.
|
Actually there's great deal of corelation.
Ever hear of body language?
It may be easy to notice that someone is worried, or happy, etc. Of
course body language cannot express our thougts in details, but all the
time we make some noises and movements. We might not be aware of it, but
there they are, sounds of our thoughts:)
| Quote: | The result of this fact is that the brain builds a model of this data by
indicating no association between private thoughts, and the physical events
represented in the sensory data. The result of that is very simple - it
leaves us with a model of realty where thoughts are disconnected from all
things physical.
This model the brain builds based on the signals it has access to, is the
source of all the confusion about human consciousness. It's why people
think the mind and the brain are two different things. People think that,
because that's exactly the model of reality the brain builds to describe
itself.
|
Well maybe. But IMHO it's much simplier than that - now that we have
intelligent systems, we need something to make distcion.
| Quote: | The net result of this is that "consciousness" is the illusion of
separation between our thoughts and our physical body. It is the source of
the illusion that humans have a soul which is separate from the body and
the reason the mind body problem has been debated for hundreds of years.
|
Well, there may be phisical reasons to belief soul is separate from
body, you know 'astral travels' and such.
But, historically... I don't think so. I've traced this view back to
Thales. He actually explained line of thought that lead him to god and
soul, goes like, how come magnets move if they're not alive? Then there
must be some universal spiritus movens...
It was Aristotle who later defined soul, seems as a need to define life:)
Anyway, before them, well there was some stories of afterlife etc, but
it was all very physical and fleshy:) Say there's no souls in
Gilgamesh... nor in Tao Te Ching:)
So, in short, seems that the distinction started around 2.5K years ago
in Mediterran, on carefully thinked philosophical basis.
| Quote: | When we look at the world around us, we believe we see the world as it is.
Rocks are hard, and pillows are soft because this is the way those things
actually are. But in fact, rocks are hard and pillows are soft because
that's the way the brain has modeled them for us.
.... |
:))
It's funny how you speak of brain as a distinct subsystem:)
When you do so, you obviously identify with something else, not brain.
Well, IMHO, *that* is consciousness. And reason to make the distinction.
Ability to think of self as an object. To think. It's on much higher
level than such a simple illusion you talk about. And furthermore, to do
so, it is necessary to separate from body/brain:)
Regards... |
|
| |
|
| Back to top |
Josip Almasi
Guest
|
| Posted: Thu Nov 20, 2008 6:55 pm Post subject: Re: Critique #1 Re: A Critique of Prof. Hubert Dreyfus' "Why |
|
|
Publius wrote:
| Quote: |
So we have a *conception* of self, not a perception.
|
Ah good point. And actually consistent with what shrinks say, like, ego
as picture of self. Or - model of self:)
To make this discussion at least a bit ontopic for c.a.nn - I've read a
book named 'neural networks with cognitive abilites', never translated
to english, where author claims such a thing can be implemented with a
supernetwork orthogonal to other (motoric etc) networks. Schematics
included:)
Note orthogonal - it's not enough to track subnet outputs, it has to
track states too.
Sounds reasonable, but such a 'center for ego' in humans is not yet found.
Regards... |
|
| |
|
| Back to top |
Isaac
Guest
|
| Posted: Thu Nov 20, 2008 7:16 pm Post subject: Re: A Critique of Prof. Hubert Dreyfus' "Why Heideggerian AI |
|
|
"Curt Welch" <curt@kcwc.com> wrote in message
news:20081119002253.348$D7@newsreader.com...
| Quote: | "Isaac" <groups@sonic.net> wrote:
"Publius" <m.publius@nospam.comcast.net> wrote in message
news:Xns9B57F131CB7B0mpubliusnospamcomcas@69.16.185.250...
"Isaac" <groups@sonic.net> wrote in
news:491f9f87$0$33506$742ec2ed@news.sonic.net:
I defy you to contrive a definition of Intelligence that works. For
example, using your current definition above, the Earth would be
intelligent because it is a system with the capacity to generate
solutions (e.g., extremely complex, yet stable atmospheric weather, ocean
currents, etc.) to solve novel problems of, for example, maintaining a
stable global temperature in the face of many (thousands) changing
(novel) variables that are constant obstacles preventing the Earth (Gia?)
from attaining her goal of minimizing temperature differences globally.
The earth is intelligent. So is the universe as a whole.
Life looks like it was designed by intelligence because
it was.
|
The problem with going this expansive on "intelligence" is that it is not
scientifically useful. For example, with your definition of intelligence a
crystal (esp. while growing) is intelligent. Can you set forth a definition
of intelligence that is scientifically useful and not just philosophically
pleasing?
| Quote: | The process of evolution is just one more example of the many
intelligent processes at work in the universe. Evolution is an example of
a reinforcement learning process
|
Evolution is not really a pure reinforcement learning process. The fitness
function select viable populations without directly reinforcing any
weightings of the genes. A selection scheme is not a reinforcement scheme.
| Quote: | and I basically consider all reinforcement
learning processes to be examples of intelligence.
|
That is far too broad. Under this scheme a tape recorder, which learns your
voice by reinforcing magnetic monopoles with your voice signal until the
signal to noise ratio reproduces your voice adequately. So, is a tape
recorder intelligent?
| Quote: |
There are many similar examples that use your language but are not
considered to be intelligent to anyone reasonable in science. Care to
update your definition or defend it?
Many people in science have no clue what they are talking about when they
use the word "intelligence". As such, they define what is, and what isn't
intelligent based on total nonsense and ungrounded speculation - as I've
said before - without using any empirical evidence to argue from.
|
Your solution goes so broad that it is not useful because computers right
now would be intelligent according to your definition, and we know, of
course, they are quite dumb (esp. Microsoft apps  . Please clarify.
| Quote: | Of course that doesn't stop them, because they like to claim things such
as
"subjective experience is outside the scope of empirical evidence".
And then they tell us what _their_ subjective experience is like and use
their beliefs about their own subjective experience to "prove" an endless
list of nonsense ideas about the universe.
The typical argument and thought path starts with the belief that human
consciousness is something that exists only in humans. Then from there,
they make the argument that since humans have this magical attribute
called
consciousness and other things like the Earth doesn't, that intelligence
requires consciousness. But since they don't have any clue what creates
human consciousness, they also don't have any clue what creates
intelligence and don't really have any way to determine if the earth is
intelligent or not.
|
true, but science always has to operate on a best working theory. Yours is
too broad and theirs is too narrow.
| Quote: |
And when asked to explain what evidence they have to suggest this
attribute
exists only in humans, they use the self serving argument that since they
"known" it exists in them, and that other humans are physically similar to
them, that this stuff they known exists in them must also exist in others.
But all that argument and the arguments that grow from it are based on a
belief that has no support. The belief that "consciousness" is something
other than simple brain function. That consciousness is not an identity
with physical brain function.
However, all the empirical evidence we have tells us that assumption is
wrong. And if we choose to believe what the empirical evidence shows us
(materialism) - then we know that there is nothing here to explain, other
than the physical signal processing that happens in the brain which
produces human behavior.
Once you grasp the significance of what the empirical evidence is telling
us, all the need of defining intelligence as some sort of link with "being
conscious" goes away. We are left with defining intelligence is some
class
of signal processing algorithm that describes how the brain works. And
though there are multiple options there, none of them make intelligence
hard to understand. It's no harder to understand than any typical machine
learning algorithm for example.
I choose to use the fairly broad and generic definition of intelligence
|
yes, too broad to be useful for anything but metaphysics.
| Quote: | being a reinforcement learning system which allows the concept to include
many processes other than just what the brain does - such as the process
of
evolution.
|
evolution is just an environmental and social selection process of who lives
and dies based on who fits an arbitrary criterion. How is that intelligent?
| Quote: | You could easily restrict the definition to something closer to what he
brain does, which would be something more like a real time distributed
parallel signal processing network trained by reinforcement instead of the
far broader "all reinforcement learning processes" I like to use.
reinforcement learning does not even begin to address the intelligence |
problem because it is method of weighting certain nodes in a network more or
less than others, but does not at all address any system level architecture
or algorithm, thus is does not provide a model of reality, just a way to
reinforce a model if you have one. For example, a classic back propagation
neural network (NN) is a implementing reinforcement learning in a NN
architecture with a propagation training algorithm; however, NN have proven
to be completely useless to do anything intelligent and cannot even be made
to converge in a hierarchical configuration even with massive (impractical)
amounts of tagged training data.
<snip>
| Quote: | Well, I think "goal" is the wrong way to understand the operation of the
brain though it's not too far off.
The true goal of a reinforcement learning machine is to maximize expected
future reward. So it's a reward maximizing machine with one prime goal.
|
greedy algorithms are only good for problems with smooth gradient descent;
however, they get trapped at local maxima and minima. Intelligence is
really looking for global maximas so reinforcement learning is a rather dumb
scheme to achieve this. Of course, GA's may try to chaotically hill hop
until a higher hill top is found, but GA's are useless because a suitable
fitness function and gene configuration space are almost impossible to
define; esp., from a top down approach.
| Quote: | What the prime goal translates into is some internal systems of values for
all possible behaviors which in turn translates into some behavior
probability distribution. This in turn must drive whatever mechanism is
in
place to select behaviors. The system that decides what behavior to
select
for the current context is using the internal system of values to pick
between alternatives.
Our higher level ideas of "goal seeking" is simply the fall out of a the
lower level behavior selection system picking the best behaviors for any
given context.
|
too vague. See above. A mountain stream searching for the best (i.e., least
energy way) to get to the bottom is implementing all the intelligent
features you propose (e.g., reinforcement of paths that work, "goal
seeking", and selecting locally optimal behaviors), however, saying that a
river defines intelligence is completely useless. Care to be more practical
and specific?
| Quote: | When you translate the implementation of the system into a reward
trained
behavior selection system, the frame problem doesn't even make much sense
to talk about. The frame problem arises nearly as much out of incorrectly
framing the question of what the purpose of the agent is. However, the
issues that surround the frame problem are real. But they are all
answered
in the context of a system which has the power to prioritize all possible
responses to stimulus signals. That is, which reaction the system chooses
at any point in time based on its learned values (priorities if you like)
is the answer to how the system deals with the frame problem. That is,
the
one problem it must solve (how to select which behavior to use at any
instant in time) is the same answer to the frame problem.
Finding a workable implementation of such a system is the path to solving
AI.
|
This is one aspect of the path to AI in the most general sense, but your
ideas are too vague (e.g., you assume an internal model of reality exist to
even be able to perceive what a frame or context is- that does not exist).
| Quote: | You said above that goals do not drive perception. That's just not true
in
my view. II think our perception and our behavior selection are one and
the
same problem. Perception is a problem of behavior selection.
|
I disagree; e.g., how is perceiving the sound of a drum to be a drum a
"problem of behavior selection"?
| Quote: | On the sensory input side of the network, the major function is
perception,
but as the signals flow through the network, the function transforms into
behavior selection. So near the output side of the network, it's mostly
"goal driven" an on the input side it's mostly "perception driven" but I
believe it's a fairly even continuum though the network as raw sensory
data
is translated to raw effector output data.
We see how this works when we test color perception of people raised (aka
trained) in different cultures with different words for different ranges
of
colors. Our perception of color bends to correctly fit the classification
of light frequency labeled by the words of our language.
of course we project our bias on what we perceive; however, that does not |
prove that perception is only based on our goal bias, which seems way off
mark to me. Please clarify.
| Quote: |
Attention is paid only
to world states which bear on the system's goals (as a background
process).
Of course, goals to play an important role in how to focus attention, and
to some extent this colors the frame problem, but I do not see how it
drives it exclusively as you put it.
It drives it exclusively in my view because behavior selection is all the
brain is doing and behavior selection works by picking behaviors that are
estimated to produce maximal expected return for the given context. And
this general process of selecting the "best" behaviors at the lowest level
is both the mechanism which creates what we think of as goal seeking and
the behavior which is think of as attention focus. I see them as one and
the same process at the low level.
the brain models phenomenon far before it forms a goal in relation to that |
phenomenon. No doubt they may be organically comingled, but to say your
goal defines your model of phenomenon would amount to a real LSD psychedelic
experience of reality, which the rest of us do not experience. :)
<snip>
| Quote: | This is because this person has never had experience with this type of
"frame" in the past and has very little experience with how to correctly
react to this combination of stimulus signals. None the less, the brain
will still pick a reaction to the stimulus signal based on the past
experience the person has had. For this guy, the "reaction" to the frame
might be to move the eyes to focus on the tree in the background because
all the city street stuff in the foreground looks mostly like "noise" to
him.
|
you are mixing high, system level goals (I want to find some food to eat)
with low level goals (like if you hear an unexpected sound turn your head
towards the sound). It is certainly not useful to say that low level goals
defines our perceptions; of course, they do filter it by pre-selecting what
our higher level systems can become aware of.
<snip>
| Quote: | I don't think anyone would say that classic AI would not return to the
world to gather more facts to add to its "millions of facts". The issue
that Dreyfus says is the problem with AI is that it creates rules that
are representations (or symbols) and are compartmentalized, both of which
he says the Philosopher Heidegger espouses, which Dreyfus and his set of
philosophers/researcher say is not the case. I think every Intelligent
system will end up effectively having a constantly evolving set of
millions of "rules", so that is not the question. Do you have any
counter examples?
Cheers!
Ariel B.-
I don't fully understand what you are suggesting here because I don't tend
to read or study the work of the type of people you are studying. I'm not
sure for example what the debate is on representations.
|
representations require a modular, hierarchical architecture that parse
realty according to the modules and objectify actions and phenomenons as
being completely separate. Dreyfus and the source he relies upon say the
brain is a flat network with no modules and have no separation between
action and phenomenons.
| Quote: |
However, just because the GOFAI approach ran into a wall after awhile,
that
doesn't mean that using "symbols" to represent something was wrong.
Dreyfus says using "symbols" in any way is very wrong. |
| Quote: | The brain uses a common language of symbols to represent everything as
well. Those symbols however are spikes. Digital computers use 1 and 0
symbols to represent things. Manipulating symbols which are
representations is exactly what the brain is doing. Any argument to the
contrary is misguided.
|
OK, explain, then, how neural networks (NN) have explicitly symbols as you
say is mandatory in an intelligent system.
| Quote: | The solution to creating human like behavior in a machine is to build
symbol manipulating machines (aka signal processors), but the symbols must
be a level closer to spikes or bits, than to English words.
I also think that the correct implementation is along the line of a
confectionist network which is processing multiple parallel signal flows.
So from that perspective, I think it's more useful to think of the network
as a signal processing machine than as "representations with symbols".
|
OK, then address the NN question above. Generally, I believe, you need a
hierarchy to represent a symbol. Dreyfus et. al. says the NN is flat.
| Quote: | But
it's the same thing no matter which way you talk about it. An AM radio
signal is still a representation of the vibration of the air, and is also
a
representation of the thoughts of the DJ which was speaking on the radio.
|
Not a good example, since the AM radio has no symbols to represent the voice
info. It is just a amplitude modulated analog signal.
| Quote: | How you choose to label these systems is matter of viewpoint far more than
a true matter of what the system is doing or how it works.
|
Not really, see my above comments on these systems you outline don't work as
you stipulate.
Thanks for your thoughtful reply!
Cheers!
Ariel-
| Quote: |
--
Curt Welch
http://CurtWelch.Com/
curt@kcwc.com
http://NewsReader.Com/ |
|
|
| |
|
| Back to top |
Publius
Guest
|
| Posted: Thu Nov 20, 2008 10:35 pm Post subject: Re: A Critique of Prof. Hubert Dreyfus' "Why Heideggerian AI |
|
|
"Isaac" <groups@sonic.net> wrote in
news:49256351$0$95529$742ec2ed@news.sonic.net:
| Quote: | The earth is intelligent. So is the universe as a whole.
Life looks like it was designed by intelligence because
it was.
The problem with going this expansive on "intelligence" is that it is
not scientifically useful. For example, with your definition of
intelligence a crystal (esp. while growing) is intelligent. Can you
set forth a definition of intelligence that is scientifically useful
and not just philosophically pleasing?
|
That is indeed a key issue, which must be resolved if everyone involved
is not to be working on different problems and trying to communicate as
though they are all on the same page.
My own favorite definition is, "The capacity of a system to solve novel
problems," with "problem" meaning any obstable or impediment the system
must overcome to reach a goal (which indeed implies that "intelligence"
is not defined for systems which lack goals).
It also distinguishes intelligence from learning. Intelligence clearly
depends upon prior learning, but mechanisms for learning (both skill
acquisition and skill refinement) are are fairly well understood. A
system which can acquire new skills by observation and emulation, or
which can increase its proficiency with practice and feedback, will
certainly improve its capacity to solve problems. But those abilities
will not equip it to deal with novel situations.
In a recent experiment with ravens a chunk of meat was suspended from a
length of string tied to a high tree branch, the meat dangling a couple
meters off the ground. The ravens at first tried to grab the meat "on
the fly," only to find it jerked from their beaks when the string went
taut. Then one of the ravens lighted on the branch at the tie point,
reached down with its beak and grabbed the string below the branch,
pulled up a loop and tucked it under its foot. It then reached down and
grabbed another loop, also trapping it under its foot. It continued this
until all the string was coiled under its foot and the bait within
reach. It then placed the chunk of meat on the branch and proceeded to
eat it.
http://www.guardian.co.uk/science/2007/apr/29/theobserversuknewspages.uknews1
The bird solved that problem in a matter of minutes, even though it was
surely a problem situation it had never before encountered. That is
intelligence. |
|
| |
|
| Back to top |
|
