[p2p-research] is the mind a computer
samuel.rose at gmail.com
Fri Nov 13 16:23:39 CET 2009
Hey Ryan, this is turning out to be an awesome discussion for sure.
Learning much from you, anyway.
On Fri, Nov 13, 2009 at 9:50 AM, Ryan Lanham <rlanham1963 at gmail.com> wrote:
> Hi Sam,
> I am familiar with Brian Arthur's work at the Santa Fe Institute on
> emergeance and complexity...it sounds like this is the same vein. That does
> help. That whole area has been productive and very interesting.
> It is emerging as a reductionist versus chaos theory sort of argument. My
> own explorations of this area were in relationship to economics...along
> lines set by people like Brian Arthur. I think I understand the basis of
> your position now.
> The information theorists being cited by Andrew are clearly reductionists,
> so the outline of disagreement is now clearer to me.
> Andrew dismisses pretty much all prior AI in favor of these current Super
> machine theorists. Would that such a clear break could occur...perhaps it
> will. I'm unconvinced so far, but their vein is also interesting.
> It would be nice to be right-tracked by some set of thoughts. I'm in favor
> of truth...not ideas, so if truth is winning, I'm happy. As a philosophical
> pragmatist, truth, for me, is what seems most right at the moment given
> reasonable vetting...so better vetting is always the aim. Better is the
> tricky word...that takes one down all sorts of philosophy of science worm
> holes. Perspective is one of them. Clearly minds are chaotic in a Santa Fe
> Institute sense of the term. But clearly there components allow for
> reductive analysis in an information theory/math sense of the term. There
> will end up being a lot of probability theory on both sides. Algorithms are
> like political theories...great to argue about...hard to move in real terms
> toward progress.
> The emergeance agenda I am familiar with has produced relatively few viable
> tools. As such, it seems to be most useful in setting challenges to
> engineering approaches. The Machine Super Intelligence crowd seems to be
> interested in tools, but so far have only produced this thing Andrew
> mentions can AIXI which is interesting but apparently useless so far. The
> work I am much more familiar with is in areas related to computer vision and
> robotics research...vision is an area that combines a lot of both of these
> topics...and it has progressed a great deal of late...3D printing in real
> time from photos is an example of a tool that comes from this work.
> Interestingly, the brain uses competing chaotic algorithms...(my source is
> research cited by Derik Bownds recently.) But it can also be reduced.
> I have my own biases toward the reductionists. But the cosmologists also
> have interesting things to say...and interesting challenges to lay down.
> Always a fascinating (and contentious) topic.
I think if you are looking to be reductionist, then the mechanism that
you want to reduce to in complex adaptive systems is the *system*, not
it's parts. A classic trope from complexity theory is that "the system
is more than the sum of it's parts".
The whole observable system is the "mechanism".
I don't think that what Eugen and J. Andrew are talking about is
actually very far off from what I am talking about. I disagree with
the way they characterize some proposed outcomes, and I propose that
what they are hoping to see arise from "machines" is more accurately
found in complex adaptive systems and emergent structures. All of
that, plus my original point that while the brain may compute, 'tis
more than computer! :-D
> On 11/13/09, Samuel Rose <samuel.rose at gmail.com> wrote:
>> Hi Ryan,
>> On Thu, Nov 12, 2009 at 11:54 AM, Ryan Lanham <rlanham1963 at gmail.com>
>> > On 11/12/09, Samuel Rose <samuel.rose at gmail.com> wrote:
>> >> "A Complex Adaptive System (CAS) is a dynamic network of many agents
>> >> (which may represent cells, species, individuals, firms, nations)
>> >> acting in parallel, constantly acting and reacting to what the other
>> >> agents are doing. The control of a CAS tends to be highly dispersed
>> >> and decentralized.
>> > This is very difficult for me to understand because the boundary
>> > conditions
>> > for what is and isn't a system are difficult to define. Something
>> > discreet
>> > occurs when a person exists. They are part of social processes, etc.,
>> > but I
>> > know of no one who quibbles much with the idea of an
>> > individual. Nations
>> > are interconnected...as are ecosystems Discreetness is much less clear.
>> > But a person could talk about resident bacteria in a human...that humans
>> > are
>> > ecosystems, etc. I'm struggling with that...
>> > There is also a macro versus micro approach issue going on I think. At
>> > one
>> > level we are discussing differences in macro systems...a brain versus a
>> > highly complicated machine (say a supercomputer running a large
>> > program).
>> > At another level there is a question as to whether processes that cause
>> > changes to genetic material...whether those are programmatic and can be
>> > simulated...or whether the workings of a set of neurons can be
>> > simulated.
>> > At the macro scale, difference of opinion is more plausible to me.
>> > So far is it the Complex Systems' peoples' opinion, Sam, that there is
>> > something at an elemental level that makes these systems different? Or
>> > is
>> > it something that occurs as complexity arises? In other words, are
>> > there
>> > boundary cases between complex and not-complex? You were getting into
>> > this
>> > a little...
>> Complex systems theory is a post-human-centric viewpoint, meaning that
>> it focuses on the nature of the system. The properties of the system
>> are more important than whether it is human or not.
>> The core property (the moment of complexity):
>> "...what a system does by virtue of its relationship to its environment"
>> http://en.wikipedia.org/wiki/Richard_Lewontin, though not a complexity
>> theorist himself, per se, gave a thorough picture of how this works
>> with living things. He also explains why this prevents us from
>> programming genes like a computer in the same book.
>> On each scale, from micro, to macro, "...what a system does by virtue
>> of its relationship to its environment" results in multi-scalar
>> emergent behavior. What you see as a "human" is now the result of
>> millions of these evolutionary relationships (what I referred to
>> earlier as "connections"). They are emergent connections.
>> It is possible to *model* emergence with computers. People have been
>> doing this for years
>> It is NOT possible to PROGRAM emergence.
>> This means that it is possible to program complicated behavior (like
>> the behavior of machines of any type), but not complex behavior (like
>> the behavior of emergent systems like ants, stockmarkets, and
>> Emergence is the result of simple rules and functions on many scales
>> *interacting* and leading to complex behavior (not complicated
>> I hope this helps.
>> > Ryan
>> Sam Rose
>> Social Synergy
>> Tel:+1(517) 639-1552
>> Cel: +1-(517)-974-6451
>> skype: samuelrose
>> email: samuel.rose at gmail.com
>> "The universe is not required to be in perfect harmony with human
>> ambition." - Carl Sagan
> Ryan Lanham
> rlanham1963 at gmail.com
> Facebook: Ryan_Lanham
> P.O. Box 633
> Grand Cayman, KY1-1303
> Cayman Islands
> (345) 916-1712
email: samuel.rose at gmail.com
"The universe is not required to be in perfect harmony with human
ambition." - Carl Sagan
More information about the p2presearch