Dear Colleagues, I have some sympathy for Pedro's call for acceptance of a fuzzy definition for "intelligence", or perhaps a large set of operational definitions. This is familiar to me as an evolutionary biologist. We treat the concept of "fitness" exactly this way, and I think both concepts hold great heuristic value even in fuzzy form. My concern under these circumstances is that we have a sufficiently clear definition that it sustains a cogent discussion. If the definition is so fuzzy that disagreements commonly boil down to presumptive differences, then serious discussion is likely to be unproductive. I would personally find it helpful to know what the limitations are on the meaning of "intelligence", and what operational definitions are being used when individuals intend to address more narrow definitions. Is it acceptable for a single entity or action to be considered intelligent by one observer and unintelligent by another?
Regards, Guy On 11/22/10 9:01 AM, "Pedro Clemente Marijuan Fernandez" <pcmarijuan.i...@aragon.es> wrote: > Dear FIS colleagues, > > very briefly stated (ugh, no spare time, devoured by ugly application > forms!), I think that quantification as Guy demands can only occur in > some small corners of our discussion areas, but not in the fundamental > ideas, not well crafted yet. For instance, I take from a recent response > of Raquel to Stan the notion of intelligence as "the capability to > process information for the purpose of adaptation or problem solving > activities. In the case of cells, problems can be caused by the > environment, extracellular aggressions, communications, etc." Well, we > can quantify (and have already done) the portions of the signaling > system involved, their correlation with genome size, etc., but have not > developed a good conceptual integration of signaling with transcription > ---and to my taste nobody as done yet, as signaling means the > "topological governance" of an enormous gene network... I mean, > premature emphasis on quantificationmay backtrack and obfuscate on > misunderstanding the big picture. > > I understand Joseph lamentations, but do not share them, as logical > clarification of an intrinsically evolutionary phenomenon --without any > major discontinuity-- as intelligence is (at least in my view), becomes > too big or too daring an undertaking. To make better sense of the > evolutionary phenomenon of intelligence, I suggested "populational > thinking" (see msg. below). Now I ad "optimality" to the mix, meaning > the presence or better the emergence of collective principles of > optimality that guide the distributed processes by the agent populations > participating in the game (roughly, optimization principles running > within cells, nervous systems, social markets). And a third ingredient , > very subtle one, could be labeled as "doctrine of limitation". It refers > to consequences of the fundamental limitations of all participants at > whatever level to have a "complete" info on the occurring collective > game, or a "complete" processing capability. In my view, this is the > most difficult and consequential point --besides, it directly militates > against the God's view we attribute to scientific observer... we already > discussed a little bit about this in Beijing! > > best wishes > > ---Pedro > > > Guy A Hoelzer escribió: >> Pedro et al., >> >> My previous cautionary post did not get much traction in this thread, but I >> still think my point was an important one to ensure that we are all talking >> about the same thing. My point was that ³intelligence² in inherently >> subjective (in the eye of the beholder), unless we can agree on the criterion >> of performance quality. I think this is necessary if we are to jump from >> mere information processing (cascades of effects resulting from the input of >> information to a system) to a notion of ³intelligence². We could, for >> example, define human intelligence as measured by performance on an IQ test. >> We could more generally define intelligence in an evolutionary context as >> measured by the fitness effects of information processing. I am personally >> not a big fan of either of these criteria. John and Pedro seem to suggest >> using the degree of ³functionality² resulting from information processing as >> a general criterion. I am intrigued by this option, although I¹m not sure >> how functionality can be measured objectively. >> >> I wonder whether this point did not get much traction previously because >> others disagree, or just don¹t think it is important. If my point is both >> correct and important, then I think we should agree on a sufficiently general >> performance criterion for the evaluation of intelligence early in this >> thread. Is there a perspective on ³intelligence² that would contradict this >> point? >> >> Regards, >> >> Guy >> >> >> On 11/19/10 4:11 AM, "Pedro Clemente Marijuan Fernandez" >> <pcmarijuan.i...@aragon.es> wrote: >> >> Dear John and FIS colleages, >> >> I much agree (below) with the return to the biological; also Gordana and >> Raquel had already argued along these guidelines. It does not mean that >> things become very much clearer initially in the connection between >> information and intelligence, but there is room for advancement. Thus, in >> Yixin's question, "What is the precise relation between intelligence and >> information?", one of the basic aspects to explore becomes "populational >> thinking" --not much considered in AI schools (perhaps very secondarily in >> the neural networks school. >> >> In fact, in all realms of intelligence in Nature (cellular, nervous systems, >> societies), we find "populations of processing agents". In cells, it is the >> population of many millions of enzymes and proteins performing catalytic >> tasks and molecular recognition activities --any emphasis in molecular >> recognition will get short of the enormous importance this phenomenon has in >> biological organization, it is the "alpha and omega" (Shu-Kun-Lin has >> produced one of the best approaches to the generality of this phenomenon). >> How populations of enzymes achieve an emergent capability of intelligence? >> Unfortunately, we can barely answer... (some googling about the term >> "cellular intelligence" will show). The discussion on neuronal intelligence >> carries a similar problem, as the neurodynamic underpinnings of animal >> behavior and animal intelligence still lack a "central theory" (most of the >> debate on consciousness is but an uninteresting quagmire)... Finally, a much >> debated contemporary topic related with social intelligence deals with the >> problem solving capacity of markets. A very extended conception about social >> organization hinges in the faith that the creativity of individuals coupled >> with the "invisible hand" of markets can solve all problems, climate change >> included... given the magnitude of civilization survival problems of today, >> the topic of social intelligence deserves some second thoughts. >> >> Anyhow, the above were just tidbits. Taken seriously, "populational thinking" >> can produce a new discourse in the relationship between information and >> intelligence. I keep saying what I argued during the Beijing conference, we >> need a new way of thinking. >> >> best regards >> >> ---Pedro >> >> >> This is a common situation in >> biology. In fact I have been told that some >> proteins pass through membranes through >> successive conformational changes that remove >> energy barriers to the transfer, much like the >> simple experiment reported in the article. This >> has been known for at least 15 years, I think. >> Inasmuch as there is functionality here, semiotic >> considerations may be relevant in this case. But >> not in the case in the article. Intelligence is a >> special case of the biological (so far). >> Conformational change is even more important and >> less dependent on the energetic substrate, and >> more on other conformations and their changes (e.g., in inference). >> >> >> >> >> >> The intelligent systems mainly do the same. >> >> >> >> >> Everything does the same. It is how it is done that is important. >> >> My best, >> John >> >> >> _______________________________________________ >> fis mailing list >> fis@listas.unizar.es >> https://webmail.unizar.es/cgi-bin/mailman/listinfo/fis >> >> _______________________________________________ fis mailing list fis@listas.unizar.es https://webmail.unizar.es/cgi-bin/mailman/listinfo/fis
