Here are some of my thoughts to the post ''Comments on "One Half Of A
Manifesto" by Jaron Lanier'' forwarded here by Francis Heylighen.
Excuse please the brainstormed, little structured, informal, and
somewhat redundant style of the present mail. A summary of my
points-of-view is added at the end of this email.
For nearly all points in the forwarded mail, I find the topics or themes
addressed hugely interesting and relevant, but as to the ways in which
the two debatants look at these issues, my first reaction is to wonder
what is actually the *use* of squabbling in that way on these subtle
points of ''metaphysics''.
Myself, I see philosophy as identical to down-to-earth empirical
science. Philosophy, IMO, is simply the highest ''meta''-level (= most
abstract level) of science; where ''science'' includes the science of
cybernetics, which latter IMO is simply the science of discovering what
are the mechanical processes behind how all ''life'' operates, and
simultaneously of trying to artificially create ''life''.
That is, IMO philosophy is simply the more ''speculative'' part of this
practical science, i.e. the part that, like a scouting party of an army,
feels out methodologies and new approaches. That is: philosophy is
maybe abstract and speculative, but IMO its ultimate *goal* is
practical, and firmly embedded in down-to-earth, practical, pragmatic,
I mean ''pragmatic technology'' as follows: a technical design that
works, works. Irrespectively of whether the design is theoretically or
esthetically ''elegant'' or not. (E.g., if we can create/design an
artificial life form that has an ugly design but which is effective at
survival, then good !)
> Date: Mon, 06 Nov 2000 22:58:15 +0100 > From: remi sussan <email@example.com> > X-Accept-Language: en,zh-CN > To: firstname.lastname@example.org > Subject: Re: philosophy: Philosophy: "One Half Of A Manifesto"
> Some of the ideas developed here have been largely exposed by Lanier > those last years, especially his rant against artificial intelligence > and intelligent agents ("intelligent agents work by making us stupid").
I don't see why considerations about things like whether intelligent agents making humans look stupid, matter. If we maybe at some point in the future succeed in creating intelligent entities/agents, then we'll see what happens. Anyway, HUMANS are *themselves* already ''intelligent agents''. I don't see how the possibility of a non-human, artificial, intelligent life form arising changes the picture -- this artificial life form will IMO arise gradually. At the point where this AI has become as intelligent as humans, the AI will become a part of human society as much as any human. An AI arising that is as intelligent as current humans will simply add one more human to human society. Humans themselves are mechanical things, as much as any future intelligent AI we will make in future. I mean: the difference between humans and AIs will simply *disappear*. Therefore, dwelling on ''who makes who look stupid'' are IMO entirely useless and extraneous.
> Not that it is the "ultimate truth", but it certainly the "state of the > art" of any discourse about reality and especially consciousness.
I think that this thought that specifically in cybernetics, this particular theory/paradigm is not the "ultimate truth", is useless. IMO, the whole approach of everything in philosophy and science is that the "ultimate truth" doesn't very much exist; and that instead of believing that science/philosophy will on some day discover The Ultimate Truth, the picture is very much more that science/philosophy proceed by ''stepwise refinement''. Every time, we proceed a little more towards better knowledge and better understanding on these few things that we have put effort into to try to discover -- and that more modest goal is enough.
IMO, this latter ''stepwise refinement'' view on science/philosophy is entirely isomorphic with how life forms evolve (''Darwinistically'':-)). Life forms evolve towards greater perfection (= greater survival effectiveness within their surroundings), -- but IMO any ''final/ultimate'' form of any life form (species) is absent.
> > 3) That subjective experience either doesn't exist, or is unimportant > > because it is some sort of ambient or peripheral effect. > > Of course, the cybernetic paradigm doesn't imply necessarily that: for > instance Varela, Thomson and Baush wrote a whole book using cybernetics > as a possible continuation of buddhist philosophy and Husserlian > phenomenology, putting the personal experience in the center of any > valuable philosophy of consciousness. An other example is Bateson > unfinished masterpiece, "Angels fear" which outlines an "epistemology of > the sacred", or Henri Atlan's "enlightenment to enlightenment".
Reading this, I get the impression that this is a repeat of the eternal war between ''materialists'' and ''idealists''. IMO, the views of idealists and materalists are above all that -- *views*, i.e., points-of-view. I don't much see the purpose of trying to prove that either point-of-view is ''true'' and the other ''false''.
I believe much more in pragmatically making use of a whole set of alternative points-of-view, in a similar pragamatic way perhaps as the Japanese simultaneously use Buddhist and Shintoist religions. They use Buddhist rituals in situations where Buddhist rituals pay off more, and use Shintoist rituals in situations where those pay off more. IMO, this ''spineless'' and pragmatic approach is entirely consistent with science, especially the more technical sciences.
In the technical sciences, it's entirely usual to use a set of incompatible models in the same field of application. In electronics e.g, in the first crude design phases of an electronic circuit, you use a very simple model of a transistor as a few simple analytical network elements. Then, when the crude design is decided on, you check the circuit with more accurate models, proably modelling also e.g. detailed effects in how charge carriers geometrically are distributed inside the transistors themselves. Never is it a problem that these models that people use in different phases of the design process, are not identical. People just use a model X if that model X is *useful* in a certain situation, and depending on what are the goals (accuracy, insight, other things) at that specific situation. That is: models are used not because they are ''True'' but because they are pragmatically *useful* in definite practical situations.
IMO, scientific and philosophical models and theories are entirely comparable to tools which e.g. a carpenter or cabinetmaker uses. Philosophy is the science of tool-making for the tools (methods) that are used in science. A carpenter is entirely not concerned with whether a tool he uses is theoretically ''True'' -- he's only interested in whether the tool is useful to him in a specific practical situation. Scientific methods/models are tools that are useful because, and to the degree that, they help give scientists insights and help them create new designs and ideas which are practically useful. The carpenter also has a whole array of different tools, which he uses pragmatically to attack different sorts of tasks. He has not one single unified tool to do all tasks. Philosophy, as the tool-maker, IMO aims to provide science with an appropriate array of tools. I mean: the goal in philosophy to seek for the one ''correct'' or ''True'' metaphysical view is entirely extraneous. If philosophy can provide science with two (or more) sorts of paradigms, A, B, ..., where each of these sheds another view on things and yields its own kinds of practically useful insights, then that's IMO excellent. It IMO is not needed that all these paradigms A, B, ..., spring from the same unified meta-paradigm or meta-model.
A certain *diversity* in the tool set of the paradigms which one uses, IMO could even potentially/possibly be more ''healthy'' than insisting on ''unifying'' all these paradigms one uses. This might in some way parallel the situation that diversity in a gene pool can enhance the survival chances of a species.
So what I want to say is that simply using in *parallel* various tastes of philosophical points-of-view, including e.g. more ''idealistic'' and more ''materialistic'' views, is IMO more useful than quibbling about how much theoretically ''True'' each of these points-of-view is, and trying to use exclusively only *one* point-of-view to the exclusion of all others. The latter IMO is similar to willfully put on blinkers (= those things that are put on horses to prevent them looking sideways).
The usefuless of a tool-maker's products is decided upon by practice, not by how much the tool-maker believes that the latest tool he has designed is more ''True'' than all other tools. In my opinion, it'd be e.g. more worthwhile to map out the areas of application where all our tools are still and are no longer very useful, and in connection with that to look more at grey areas between different models (e.g. see where two different approaches applied to the same situation yield the same final result, then analyze why).
> > 4) That what Darwin described in biology, or something like it, is in fact > > also the singular, superior description of all creativity and culture. > > What is true for cybernetics is true for darwinism. One can have a crude > or subtle interpretation of darwinism. Accepting darwinism as a worthy > paradigm for all kinds of evolution doesn't mean that we must accept all > the simplest versions of darwinism.
IMO, crude models have their usefulness as well as more detailed models. In some situations, it is not very useful to insist on applying the most detailed, cumbersome model that one has. The very ancient geometry theses of Euclides are still very useful, as are the laws of Newton. An architect and a pilot of a mars lander don't use the latest, newest, most ''state-of-the-art'' scientific models like non-Euclidic geometry and string theory. The old models are very very often the basis on which newer theories are built -- the newer theories in that case don't disprove the old ones, they merely fill in additional accuracy in some corners.
> > 5) That qualitative as well as quantitative aspects of information systems > > will be accelerated by Moore's Law. > > I certainly agree with that. there is nothing more irritating, when you > try to discuss artificial intelligence, to hear that its feasibility is > dependent of the Moore law, as if better processor, better RAM and > better hard disks were sufficient to reproduce consciousness. But again, > this is not an argument against A.I. per se . Only a critic of their > more naive aficionados.
> > 6) That biology and physics will merge with computer science (becoming > > biotechnology and nanotechnology), resulting in life and the physical > > universe becoming mercurial; [...] > > There is certainly a mythic component in this vision, and the similarity > between this technological singularity and Terence McKenna eschaton for > instance, speaks for itself. Erik Davis wrote a lot of interesting pages > about it in Techgnosis. BUT again, this mythical aspect doesn't impair > that the merging of computer science, biology and physics have really > solid basis which cannot be easily dismissed because they are mythic. In > other words, that's not because Icarus was a myth that airplanes cannot > exist.
What I don't understand in these considerations is that people apparently don't accept that HUMANS themselves must (somehow) already be mechanisms, automatons. I fail to see how any cybernetician could believe otherwise.
IMO, since humans are already mechanical automatons, quibbling about the question whether ''biology and physics will merge with computer science (becoming biotechnology and nanotechnology)'' is extraneous. It's like two intelligent computers discussing whether it is possible to construct a computer.
I fail entirely to see how labelling things as ''mythical'', and/or fighting about whether it's good or bad to look at things in lights in which things or goal appear as ''mythical'', could be helpful.
For the purpose of helping to pry loose additional and (hopefully:-)) useful discussions on these topics discussed in the mail forwarded by Francis Heylighen, I propose an additional way of looking at these topics, which is in a way ''orthogonal'' to the paths of thought that *both* autors of the forwarded mail use in looking at things. As follows:
- Alternative theories can be used simultaneously, just like a carpenter uses a whole array of tools simultaneously, where he knows the situations in which each individual tool is, and is not, useful. The carpenter is not very much looking for one unified tool with which he can do everything.
- Each theory is a point-of-view, that can be useful in some situations (and can be less useful in others). Each theory has its specific areas of application, which are identical to the areas where the theory still yields useful insights. The fact that a specific theory X is not useful in some sub-set of situations does not mean that one has to entirely throw away that tool.
- The usefulness of a theory can be compared to the effectiveness at survival of a biological species.
- Successive refinement of theories can be compared to evolutionary change/betterment of a biological species.
- The merit of a point-of-view or model has to be proved in *practice*, e.g. by whether the model leads to the possibility of creating new or better forms/kinds of AI in practice. It's useless to dwell on entirely *theoretical* merits (such as ''truth'') of a model/point-of-view. This parallels the situation that in biological evolution, the survival effectiveness of a species must be proven in *practice*, in the everyday struggle for survival of the species.
Well, I hope that the above was not entirely un-useful. :-)
Best regards, Menno (email@example.com)
Ir. Menno Rubingh, Scientific programmer, Software designer, & Software documentation writer Doelenstraat 62, 2611 NV Delft, Netherlands phone +31 15 2146915 (answering machine backup) email firstname.lastname@example.org http://www.rubinghscience.org/ ======================================== Posting to email@example.com from "Menno RUBINGH" <firstname.lastname@example.org>
This archive was generated by hypermail 2b29 : Tue Nov 14 2000 - 02:28:17 GMT