Loading ...
Sorry, an error occurred while loading the content.
 

Re: [nanotech] on nanotubes

Expand Messages
  • Mark Gubrud
    Stuart Hameroff is a professor of anesthesiology at Arizona State. His 1988 book Ultimate Computing talked a bit about nanotechnology and advanced the
    Message 1 of 13 , Sep 7, 2002
      Stuart Hameroff is a professor of anesthesiology at Arizona State. His
      1988 book "Ultimate Computing" talked a bit about nanotechnology and
      advanced the unlikely theory that the microtubules (not nanotubes)
      forming the cytoskeleton found inside brain cells were somehow the "seat
      of consciousness," based on their supposed ability to function as
      cellular automata, a hot buzzword at the time, and hence as universal
      computers... Later, Hameroff updated this to the not merely unlikely,
      but utterly preposterous theory that the microtubules actually function
      as quantum computers... A suggestion that was decisively debunked by Max
      Tegmarck (apparently the first person to realize that such an obvious
      point was nevertheless worth writing a paper on); See
      http://www.hep.upenn.edu/~max/brain1.html

      Microtubules may indeed have cellular automata properties and certainly
      are involved in intracellular communication processes but their
      computations, if any, are of a primitive, local and inexact sort; they
      are not universal computers and certainly not quantum computers.

      Hameroff teams up with Roger Penrose, who denies the possibility of
      artificial intelligence based on some funky logic arguments, to advance
      the mystical view that "consciousness" is somehow connected with quantum
      mechanics and can't be understood in any terms less than the
      still-unavailable "correct quantum theory of gravity."

      For those interested in some fine examples of tenured and peer-reviewed
      pseudoscience, see
      www.consciousness.arizona.edu/hameroff/

      > paul wrote:
      >
      > is this related to Hameroff's work on the tub properities of
      > cytoskeleton as in Hammeroff's book "Ultimate Computing".?
    • Eliezer S. Yudkowsky
      ... It pays not to underestimate the cleverness of the brain. Of course, it pays not to overestimate it either. But neurons are certainly not the simple toys
      Message 2 of 13 , Sep 8, 2002
        Mark Gubrud wrote:
        >
        > Microtubules may indeed have cellular automata properties and certainly
        > are involved in intracellular communication processes but their
        > computations, if any, are of a primitive, local and inexact sort; they
        > are not universal computers and certainly not quantum computers.

        It pays not to underestimate the cleverness of the brain. Of course, it
        pays not to overestimate it either. But neurons are certainly not the
        simple toys that they appear to be in most so-called "neural networks",
        and many attempts at realistic neural simulation are forced to descend to
        the microtubular level, if I recall correctly. I would certainly not
        characterize the hypothesis of computational roles for microtubules as
        Hameroff's particular baby, regardless of whether that hypothesis happens
        to be necessary to allow for the noncomputational brain processes
        advocated by Penrose.

        --
        Eliezer S. Yudkowsky http://singinst.org/
        Research Fellow, Singularity Institute for Artificial Intelligence
      • Mark Gubrud
        Microtubules are involved in intracellular transport of proteins and mRNAs, and in neurons that role is probably computational in the sense of what you might
        Message 3 of 13 , Sep 8, 2002
          Microtubules are involved in intracellular transport of proteins and
          mRNAs, and in neurons that role is probably "computational" in the sense
          of what you might call "learning algorithms" that the neuron uses to
          govern growth and tuning of synapses and membrane processes, but no way
          do the microtubules act as exact cellular automata, much less implement
          any Turinglike universal computer, as was suggested by Hameroff, but
          which is entirely inconsistent with biology in general. These days we
          have a more expansive view of what constitutes "computation" and I
          entirely agree the neuron cannot be reduced to anything like the
          simplistic models used in artificial neural nets, but bear in mind that
          the timescale for significant transfers of information (i.e. molecules)
          along microtubules is much slower than that for impulse conduction by
          membranes and neurotranmission across synapses, so it wouldn't make much
          sense to suppose that such microtubule "computation" explains the
          phenomenon of consciousness; rather, it can be considered as part of the
          internal machinery that supports learning and memory, whereas the
          realtime "computation" involved in conscious activity mostly occurs in
          the membrane and membrane-bound protein machinery. Any computation
          which is taking place at the molecular level is going to be very
          low-level and cannot plausibly be highly coded in terms of local
          spatiotemporal patterns of activity; rather its significance is coded by
          its location within the global system.

          "Eliezer S. Yudkowsky" wrote:
          >
          > Mark Gubrud wrote:
          > >
          > > Microtubules may indeed have cellular automata properties and certainly
          > > are involved in intracellular communication processes but their
          > > computations, if any, are of a primitive, local and inexact sort; they
          > > are not universal computers and certainly not quantum computers.
          >
          > It pays not to underestimate the cleverness of the brain. Of course, it
          > pays not to overestimate it either. But neurons are certainly not the
          > simple toys that they appear to be in most so-called "neural networks",
          > and many attempts at realistic neural simulation are forced to descend to
          > the microtubular level, if I recall correctly. I would certainly not
          > characterize the hypothesis of computational roles for microtubules as
          > Hameroff's particular baby, regardless of whether that hypothesis happens
          > to be necessary to allow for the noncomputational brain processes
          > advocated by Penrose.
        • paul
          Mark Gubrud University of Maryland, (Open letter, in response to a discussion in the nanotech e-forum), Mark, You make a number of statements as if axioms, and
          Message 4 of 13 , Sep 8, 2002
            Mark Gubrud
            University of Maryland,
             
            (Open letter, in response to a discussion in the nanotech e-forum),
             
            Mark,
             
            You make a number of statements as if axioms, and have some argument that is pointed to as the proof.    The argument is well made (included below for the bcc list), except the degree of authority that is assumed by your self and by Max Tegmark (as cited).   Many of us have looked at this viewpoint.   But some of "us" believe that this is an open question. 
             
            I am opening this up to a number of other groups in a bcc, and some individual scholars may decide to join the discussion at nanotech.  I hope that IF this happens that we will all be respectful while making our viewpoint specific.  Gina Miller has proved herself an excellent moderator.  Stu Hameroff's home page is 
             
             
            and one should look closely as to why Stuart feels that anesthesiology sheds an important light on consciousness. 
             
            For me, since meeting him in 1992, I have felt that the issue was about the mechanisms involved in emergence, at various levels of physical-energy organization (this became later for me "stratified complexity theory".) 
             
            For reductionist science to gets it's way in nano systems (as it has in Artificial Intelligence)  would be a big mistake to the science, because most of what we see in a nano system is the consequence of our looking or some other induction of emergence.  (Perhaps?)
             
            Perhaps a vetting of this issue might be developed as part of that more focused topic in nanotech@yahoogroups.com
             
            This type of on-line discussions are very hard to moderate because participants do not take the time, very often, to see what the "other" is saying.  And this is a very complex discussion, anytime one is talking about nano events, particularly if the nano event is involved in either being the result of emergence, or resulting in emergence. 
             
            I developed some mathematics on generalized I-sing (planer rotators) { J. Kowalski; A. Ansari; P. Prueitt; R. Dawes and G. Gross (1988.) On Synchronization and Phase Locking in Strongly Coupled Systems of Planar Rotators. Complex Systems 2, 441-462. } model and used this to conjecture a Process Compartment Hypothesis (PCH) that looks that this issue of emergence within a theory of stratification.
             
             
             
            for me, this mathematics (and some un-published extensions of this PCH) has been my guide.  In the mathematics in Chapter 1 of my book, one sees the nature of the problem of using a classical (first order differential equation) as a model of anything that has emergence as part of the natural phenomenon.  Thus leading to the study of the work of Robert Rosen (theoretical biology and category theory) and then the work of Peter Kugler (perceptional measurement). 
             
             
            This was in the context of evaluating the Russian work in information warfare and specifically situational logics.
             
            Please let us recognize that "emergence" is not accounted for at all in Newtonian physics or in any variation of logical positivism or reductionism.  If there is an "accounting" of "emergence" within these paradigms, then they are controversial. 
             
            There are some of us that think that Penrose and Hameroff are moving down a path that provides an explanation of consciousness.  John Eccles also has work on the synapse mechanics for delay of the process involved, however this process is made up.    On the specifics of Penrose - Hameroff self orchestrated collapse, we have unfinished work.
             
             
            However, this unfinished work can defend itself against popular science based on reductionism - particularly the type of pronouncements that Churchland and others advance form a part of the Neural Network community.
             
            In my mind, this "work in progress" fits well with what I understand of Pribram's work on holonomic theory.  And then , again in my mind, this form of "scientific realism" seem to be consistent with J. J. Gibson's work and the school of ecological physics (action -perception cycles). 
             
             
             Then perhaps one might appreciate Walter Freeman's work on the brain as a sensory processor in which reaction-cause cycles do not exist ONLY at one level of organization, but at many levels of organization (between 3 and 17). 
             
             
            Is some work that I have done on converting some non-reductionist, what we call stratified complexity, "science" into computer architecture.  I realize that I have made many errors, but the argument is made that we are addressing open questions tentatively (conjecturally) as an attempt to explain what has not been explained so far. 
             
            The language that you use is all (in my mind) used to force others to agree that there is nothing to talk about.  Just accept what you say.
             
            Consciousness exists.  Moreover the "ends" of science as science currently defines itself seem always to be implicated in the investigations of the POSSIBILITY of a science of consciousness.
             
            What is often left after the destruction of  evidence regarding a quantum mechanical role in memory, awareness and anticipation is the AI Mythology. 
             
             
            and this is little more than a religion itself - has there is simply no evidence that computer programs are capable of consciousness. 
             
            Reductionism, and I think that this is what is at the heart of your (let me say) "bias", has some real questions to answer regarding it's validity.  The utility of reductionism has been proven in many ways, but its failures have also been demonstrated specifically in the failure of AI to produce predictive methodology for things like terrorism attacks. 
             
            The stratified complexity theory sees something as a terrorism attack as being a matter of scale. 
             
            The terrorism has a set of causes that are very stable and produce the potential for an event of a specific type.  Then at a different level of causal organization we have a small social group that acquires the material substance to carry out an attack.  The environment level of organization then facilitates or not the actual event.  This can be studied at each level of organization by acquiring what is in essence at system of signs (semiotics) that are interpreted by human minds.
             
             
             
            So we have the basic elements for decomposition of event structures into period tables (due to the stability of the causes of terrorism).  Of course one must include in these constant levels of cause, the imperfections of humans who are suffering and the various forms of religious fundamentalism and also the justification of scientific fundamentalism.  This need to justify this way of thinking can be an obstacle to our (social) understanding of the nature of mental and social events.
             
            The development of a complete technology for "synthetic intelligence" using human/computer interaction is outlined in these two papers:
             
             
            I ask that these be reviewed within the various communities, and that I be forgiven not making it clear that most of this is conjectural.
             
            The software that OntologyStream has produced, however, is not conjectural.
             
             
            and is available for evaluation.
             
            One more thing, the claim here is that formal systems have a specific glass ceiling that can not be seen when believing in the formal system ( as occurs in pure reductionism).  One might see this in the AI and SQL based technologies for text analysis, similarity analysis etc that we now deploy within the government (and within business).  There is even a literature on the "false sensemaking"  (Weick) that occurs (often) in acting as if these systems will provide what they can not provide (i.e., Predictive Analysis Methodology). 
             
            It is vital, to our Nation, to develop and deploying something like Synthetic Intelligence as part of the War effort.  paul@...
             
             
            I am a scientist, not a business person; so please understand my need to communicate freely regarding this issue.  I do understand that my work in controversial, I just claim that some of it is correct.
             
             
            Dr. Paul S. Prueitt
            Founding Committee BCNGroup (1991)
            CEO OntologyStream Inc.
            Chantilly VA
            703-981-2676
             
             
            -----Original Message-----
            From: Mark Gubrud [mailto:mgubrud@...]
            Sent: Saturday, September 07, 2002 11:52 PM
            To: nanotech@yahoogroups.com
            Subject: Re: [nanotech] on nanotubes

            Stuart Hameroff is a professor of anesthesiology at Arizona State.  His
            1988 book "Ultimate Computing" talked a bit about nanotechnology and
            advanced the unlikely theory that the microtubules (not nanotubes)
            forming the cytoskeleton found inside brain cells were somehow the "seat
            of consciousness," based on their supposed ability to function as
            cellular automata, a hot buzzword at the time, and hence as universal
            computers... Later, Hameroff updated this to the not merely unlikely,
            but utterly preposterous theory that the microtubules actually function
            as quantum computers... A suggestion that was decisively debunked by Max
            Tegmarck (apparently the first person to realize that such an obvious
            point was nevertheless worth writing a paper on); See
            http://www.hep.upenn.edu/~max/brain1.html

            Microtubules may indeed have cellular automata properties and certainly
            are involved in intracellular communication processes but their
            computations, if any, are of a primitive, local and inexact sort; they
            are not universal computers and certainly not quantum computers. 

            Hameroff teams up with Roger Penrose, who denies the possibility of
            artificial intelligence based on some funky logic arguments, to advance
            the mystical view that "consciousness" is somehow connected with quantum
            mechanics and can't be understood in any terms less than the
            still-unavailable "correct quantum theory of gravity."

            For those interested in some fine examples of tenured and peer-reviewed
            pseudoscience, see
            www.consciousness.arizona.edu/hameroff/

            > paul wrote:
            >
            > is this related to Hameroff's work on the tub properities of
            > cytoskeleton as in Hammeroff's book "Ultimate Computing".?



            The Nanotechnology Industries mailing list.
            "Nanotechnology: solutions for the future."
            www.nanoindustries.com


            Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.
          • Mark Gubrud
            I visited the web site found at the bottom of this pile of gibberish. I sure wouldn t advise anyone who isn t interested in losing money to invest in any of
            Message 5 of 13 , Sep 8, 2002
              I visited the web site found at the bottom of this pile of gibberish.
              I sure wouldn't advise anyone who isn't interested in losing money to
              invest in any of Prueitt's schemes.

              > paul wrote:
              >
              > Mark Gubrud
              > University of Maryland,
              >
              > (Open letter, in response to a discussion in the nanotech e-forum),
              >
              > Mark,
              >
              > You make a number of statements as if axioms, and have some argument
              > that is pointed to as the proof. The argument is well made
              > (included below for the bcc list), except the degree of authority that
              > is assumed by your self and by Max Tegmark (as cited). Many of us
              > have looked at this viewpoint. But some of "us" believe that this is
              > an open question.
              >
              > I am opening this up to a number of other groups in a bcc, and some
              > individual scholars may decide to join the discussion at nanotech. I
              > hope that IF this happens that we will all be respectful while making
              > our viewpoint specific. Gina Miller has proved herself an excellent
              > moderator. Stu Hameroff's home page is
              >
              > http://www.consciousness.arizona.edu/hameroff/
              >
              > and one should look closely as to why Stuart feels that anesthesiology
              > sheds an important light on consciousness.
              >
              > For me, since meeting him in 1992, I have felt that the issue was
              > about the mechanisms involved in emergence, at various levels of
              > physical-energy organization (this became later for me "stratified
              > complexity theory".)
              >
              > For reductionist science to gets it's way in nano systems (as it has
              > in Artificial Intelligence) would be a big mistake to the science,
              > because most of what we see in a nano system is the consequence of our
              > looking or some other induction of emergence. (Perhaps?)
              >
              > Perhaps a vetting of this issue might be developed as part of that
              > more focused topic in nanotech@yahoogroups.com .
              >
              > This type of on-line discussions are very hard to moderate because
              > participants do not take the time, very often, to see what the "other"
              > is saying. And this is a very complex discussion, anytime one is
              > talking about nano events, particularly if the nano event is involved
              > in either being the result of emergence, or resulting in emergence.
              >
              > I developed some mathematics on generalized I-sing (planer rotators) {
              > J. Kowalski; A. Ansari; P. Prueitt; R. Dawes and G. Gross (1988.) On
              > Synchronization and Phase Locking in Strongly Coupled Systems of
              > Planar Rotators. Complex Systems 2, 441-462. } model and used this to
              > conjecture a Process Compartment Hypothesis (PCH) that looks that this
              > issue of emergence within a theory of stratification.
              >
              > http://www.bcngroup.org/area3/pprueitt/kmbook/Chapter1.htm
              >
              >
              > for me, this mathematics (and some un-published extensions of this
              > PCH) has been my guide. In the mathematics in Chapter 1 of my book,
              > one sees the nature of the problem of using a classical (first order
              > differential equation) as a model of anything that has emergence as
              > part of the natural phenomenon. Thus leading to the study of the work
              > of Robert Rosen (theoretical biology and category theory) and then the
              > work of Peter Kugler (perceptional measurement).
              >
              >
              > This was in the context of evaluating the Russian work in information
              > warfare and specifically situational logics.
              >
              > Please let us recognize that "emergence" is not accounted for at all
              > in Newtonian physics or in any variation of logical positivism or
              > reductionism. If there is an "accounting" of "emergence" within these
              > paradigms, then they are controversial.
              >
              > There are some of us that think that Penrose and Hameroff are moving
              > down a path that provides an explanation of consciousness. John
              > Eccles also has work on the synapse mechanics for delay of the process
              > involved, however this process is made up. On the specifics of
              > Penrose - Hameroff self orchestrated collapse, we have unfinished
              > work.
              >
              > http://www.consciousness.Arizona.edu/Hameroff/Pen-Ham/More_Neural/more%20Neural%20than%20Pat.htm
              >
              > However, this unfinished work can defend itself against popular
              > science based on reductionism - particularly the type of
              > pronouncements that Churchland and others advance form a part of the
              > Neural Network community.
              >
              > In my mind, this "work in progress" fits well with what I understand
              > of Pribram's work on holonomic theory. And then , again in my mind,
              > this form of "scientific realism" seem to be consistent with J. J.
              > Gibson's work and the school of ecological physics (action -perception
              > cycles).
              >
              > http://www.bcngroup.org/area3/pprueitt/kmbook/Chapter7.htm
              >
              > Then perhaps one might appreciate Walter Freeman's work on the brain
              > as a sensory processor in which reaction-cause cycles do not exist
              > ONLY at one level of organization, but at many levels of organization
              > (between 3 and 17).
              >
              > http://www.bcngroup.org/area3/pprueitt/kmbook/Chapter5.htm
              >
              > Is some work that I have done on converting some non-reductionist,
              > what we call stratified complexity, "science" into computer
              > architecture. I realize that I have made many errors, but the
              > argument is made that we are addressing open questions tentatively
              > (conjecturally) as an attempt to explain what has not been explained
              > so far.
              >
              > The language that you use is all (in my mind) used to force others to
              > agree that there is nothing to talk about. Just accept what you say.
              >
              > Consciousness exists. Moreover the "ends" of science as science
              > currently defines itself seem always to be implicated in the
              > investigations of the POSSIBILITY of a science of consciousness.
              >
              > What is often left after the destruction of evidence regarding a
              > quantum mechanical role in memory, awareness and anticipation is the
              > AI Mythology.
              >
              > http://www.ontologystream.com/administration/DARPA/Commentary.htm
              >
              > and this is little more than a religion itself - has there is simply
              > no evidence that computer programs are capable of consciousness.
              >
              > Reductionism, and I think that this is what is at the heart of your
              > (let me say) "bias", has some real questions to answer regarding it's
              > validity. The utility of reductionism has been proven in many ways,
              > but its failures have also been demonstrated specifically in the
              > failure of AI to produce predictive methodology for things like
              > terrorism attacks.
              >
              > The stratified complexity theory sees something as a terrorism attack
              > as being a matter of scale.
              >
              > The terrorism has a set of causes that are very stable and produce the
              > potential for an event of a specific type. Then at a different level
              > of causal organization we have a small social group that acquires the
              > material substance to carry out an attack. The environment level of
              > organization then facilitates or not the actual event. This can be
              > studied at each level of organization by acquiring what is in essence
              > at system of signs (semiotics) that are interpreted by human minds.
              >
              > http://www.ontologystream.com/threeYearPlan/mentalEvent.htm
              >
              >
              > So we have the basic elements for decomposition of event structures
              > into period tables (due to the stability of the causes of terrorism).
              > Of course one must include in these constant levels of cause, the
              > imperfections of humans who are suffering and the various forms of
              > religious fundamentalism and also the justification of scientific
              > fundamentalism. This need to justify this way of thinking can be an
              > obstacle to our (social) understanding of the nature of mental and
              > social events.
              >
              > The development of a complete technology for "synthetic intelligence"
              > using human/computer interaction is outlined in these two papers:
              >
              > http://www.ontologystream.com/beads/frameworks/generalFrameworks.htm
              > http://www.ontologystream.com/threeYearPlan/mentalEvent.htm
              >
              > I ask that these be reviewed within the various communities, and that
              > I be forgiven not making it clear that most of this is conjectural.
              >
              > The software that OntologyStream has produced, however, is not
              > conjectural.
              >
              > http://www.ontologystream.com/cA/index.htm
              >
              > and is available for evaluation.
              >
              > One more thing, the claim here is that formal systems have a specific
              > glass ceiling that can not be seen when believing in the formal system
              > ( as occurs in pure reductionism). One might see this in the AI and
              > SQL based technologies for text analysis, similarity analysis etc that
              > we now deploy within the government (and within business). There is
              > even a literature on the "false sensemaking" (Weick) that occurs
              > (often) in acting as if these systems will provide what they can not
              > provide (i.e., Predictive Analysis Methodology).
              >
              > It is vital, to our Nation, to develop and deploying something like
              > Synthetic Intelligence as part of the War effort.
              > paul@...
              >
              > http://www.ontologystream.com/SLIP/index1.htm
              >
              > I am a scientist, not a business person; so please understand my need
              > to communicate freely regarding this issue. I do understand that my
              > work in controversial, I just claim that some of it is correct.
              >
              >
              > Dr. Paul S. Prueitt
              > Founding Committee BCNGroup (1991)
              > CEO OntologyStream Inc.
              > Chantilly VA
              > 703-981-2676
            • Eliezer S. Yudkowsky
              ... You use up your credibility as a skeptic when you use it indiscriminately against pseudoscience, and honestly attempted science which you happen to
              Message 6 of 13 , Sep 8, 2002
                Mark Gubrud wrote:
                > I visited the web site found at the bottom of this pile of gibberish.
                > I sure wouldn't advise anyone who isn't interested in losing money to
                > invest in any of Prueitt's schemes.

                You use up your credibility as a skeptic when you use it indiscriminately
                against pseudoscience, and honestly attempted science which you happen to
                disagree with.

                --
                Eliezer S. Yudkowsky http://singinst.org/
                Research Fellow, Singularity Institute for Artificial Intelligence
              • Mark Gubrud
                So, are you saying you disagree with me about this?
                Message 7 of 13 , Sep 8, 2002
                  So, are you saying you disagree with me about this?

                  "Eliezer S. Yudkowsky" wrote:
                  >
                  > Mark Gubrud wrote:
                  > > I visited the web site found at the bottom of this pile of gibberish.
                  > > I sure wouldn't advise anyone who isn't interested in losing money to
                  > > invest in any of Prueitt's schemes.
                  >
                  > You use up your credibility as a skeptic when you use it indiscriminately
                  > against pseudoscience, and honestly attempted science which you happen to
                  > disagree with.
                • Mark Gubrud
                  Pruiett s website, http://www.ontologystream.com/SLIP/index1.htm indicates he seeks to sell stock in his company. Based on what I see there, I wouldn t
                  Message 8 of 13 , Sep 8, 2002
                    Pruiett's website,

                    http://www.ontologystream.com/SLIP/index1.htm

                    indicates he seeks to sell stock in his company. Based on what I see
                    there, I wouldn't consider it a very sound investment. No, not at all.

                    Are you saying you disagree with my assessment, Eliezer?

                    "Eliezer S. Yudkowsky" wrote:
                    >
                    > Mark Gubrud wrote:
                    > > I visited the web site found at the bottom of this pile of gibberish.
                    > > I sure wouldn't advise anyone who isn't interested in losing money to
                    > > invest in any of Prueitt's schemes.
                    >
                    > You use up your credibility as a skeptic when you use it indiscriminately
                    > against pseudoscience, and honestly attempted science which you happen to
                    > disagree with.
                  • Eliezer S. Yudkowsky
                    ... I disagree with your assessment of Pruiett as gibberish . I don t think this is a correct approach to Artificial Intelligence, but I would not
                    Message 9 of 13 , Sep 8, 2002
                      Mark Gubrud wrote:
                      > Pruiett's website,
                      >
                      > http://www.ontologystream.com/SLIP/index1.htm
                      >
                      > indicates he seeks to sell stock in his company. Based on what I see
                      > there, I wouldn't consider it a very sound investment. No, not at all.
                      >
                      > Are you saying you disagree with my assessment, Eliezer?

                      I disagree with your assessment of Pruiett as "gibberish". I don't think
                      this is a correct approach to Artificial Intelligence, but I would not
                      characterize it as "gibberish" but rather honest but mistaken material
                      which is not at all unusual as honest but mistaken material in AI goes.
                      Similarly it is possible that Pruiett's approach to analyzing natural
                      language may work to turn up some useful regularities in text streams,
                      even though it will never work to produce real AI or even work as well as
                      Pruiett thinks it will. His corporation is not going to be the last
                      corporation humanity needs to form, but it may make a profit - though I
                      wouldn't know without looking at the approach in a lot more detail.
                      Moreover, if I were to provide that as advice to an investor (that
                      Pruiett's approach is not likely to be a major breakthrough, regardless of
                      whether it produces saleable minor useful results), I would insert the
                      disclaimer that it is advice based on a theory of intelligence which,
                      though it happens to be actually correct :), is not accepted by the
                      contemporary field as a whole.

                      Your dismissal seemed to me to have been drawn from the same class as your
                      (correct) dismissal of antigravity attempts; an implication that Pruiett
                      was a pseudoscientist and that even people looking to invest in
                      classicalist approaches to AI should stay away from his dangerous fraud.
                      In which case I have not, so far, picked up anything that would justify
                      such an implication. From scanning through the text, he looks like an
                      ordinary AI researcher to me, and if you're going to advise investors to
                      stay away from him solely on that basis you should make it clear that your
                      dismissal is directed at classicalist AI approaches as a whole rather than
                      Pruiett in particular.

                      --
                      Eliezer S. Yudkowsky http://singinst.org/
                      Research Fellow, Singularity Institute for Artificial Intelligence
                    • Mark Gubrud
                      ... Seems like a marginal case to me. I thought Pruiett s open letter addressed to me was loaded with gibberish. I can t say that I know all of the
                      Message 10 of 13 , Sep 9, 2002
                        "Eliezer S. Yudkowsky" wrote:

                        > I disagree with your assessment of Pruiett as "gibberish".....
                        >
                        > Your dismissal seemed to me to have been drawn from the same class as your
                        > (correct) dismissal of antigravity attempts; an implication that Pruiett
                        > was a pseudoscientist and that even people looking to invest in
                        > classicalist approaches to AI should stay away from his dangerous fraud.
                        > In which case I have not, so far, picked up anything that would justify
                        > such an implication. From scanning through the text, he looks like an
                        > ordinary AI researcher to me, and if you're going to advise investors to
                        > stay away from him solely on that basis you should make it clear that your
                        > dismissal is directed at classicalist AI approaches as a whole rather than
                        > Pruiett in particular.

                        Seems like a marginal case to me. I thought Pruiett's "open letter"
                        addressed to me was loaded with gibberish. I can't say that I know all
                        of the software advertised on his website is useless, but I suspect that
                        might be the case. I don't think "he looks like an ordinary AI
                        researcher," because he seems to make very free and inappropriate use of
                        buzzwords like "fractal" and "holonomic" and seems to have quite an
                        obsession with mystical, grandiose notions of "emergence" and "general
                        systems theory" and the idea that there is a mysterious difference
                        between human and artificial intelligence (I would say that there is a
                        difference, but it is not mysterious, because humans and computers are
                        completely different systems, neither one explained by any new physics
                        but both simply by their complicated structures).

                        The more you read of Pruiett's material the funnier it smells... but
                        I'll admit I haven't read that much of it. There does seem to be a core
                        of software he has developed which looks straight and might be good for
                        something, but given the surrounding atmospherics I wouldn't bet on it.

                        I'm not sure what you mean by "classicalist AI approaches" but by no
                        means do I dismiss AI as the art and science of building complicated
                        information processing systems capable of ever higher levels of
                        automation of intellectual work. The bottom line for any AI system is
                        "does it do the job?" and "how well?" - not "is it genuine
                        intelligence?"
                      • Eliezer S. Yudkowsky
                        ... Unfortunately, *this is normal* for AI researchers. At least insofar as I have been able to observe. -- Eliezer S. Yudkowsky
                        Message 11 of 13 , Sep 9, 2002
                          Mark Gubrud wrote:
                          > I don't think "he looks like an ordinary AI
                          > researcher," because he seems to make very free and inappropriate use of
                          > buzzwords like "fractal" and "holonomic" and seems to have quite an
                          > obsession with mystical, grandiose notions of "emergence" and "general
                          > systems theory" and the idea that there is a mysterious difference
                          > between human and artificial intelligence (I would say that there is a
                          > difference, but it is not mysterious, because humans and computers are
                          > completely different systems, neither one explained by any new physics
                          > but both simply by their complicated structures).

                          Unfortunately, *this is normal* for AI researchers. At least insofar as I
                          have been able to observe.

                          --
                          Eliezer S. Yudkowsky http://singinst.org/
                          Research Fellow, Singularity Institute for Artificial Intelligence
                        Your message has been successfully submitted and would be delivered to recipients shortly.